BestHub
Sign in
Tagged articles
1235 articles
Page 6 of 13
DataFunSummit
DataFunSummit
Jun 21, 2022 · Artificial Intelligence

JiuGe: An Automatic Chinese Classical Poetry Generation System – Algorithms and Research Overview

This article presents the JiuGe system developed by THUNLP for automatically generating Chinese classical poetry, detailing its research motivations, model architecture—including salient‑clue, working‑memory, topic‑memory, style‑transfer and reinforcement‑learning components—implementation, applications, and future directions.

Artificial IntelligenceDeep LearningPoetry Generation
0 likes · 18 min read
JiuGe: An Automatic Chinese Classical Poetry Generation System – Algorithms and Research Overview
DataFunTalk
DataFunTalk
Jun 15, 2022 · Artificial Intelligence

Data Interaction Based Click‑Through Rate Model (RIM): Review, Architecture, and Experimental Insights

This article reviews the evolution of click‑through rate (CTR) prediction models from early logistic regression and factorization machines to deep neural networks, introduces the data‑interaction based RIM (Retrieval & Interaction Machine) architecture with its search and prediction modules, and presents extensive experimental comparisons and future research directions.

CTRDeep LearningRIM
0 likes · 14 min read
Data Interaction Based Click‑Through Rate Model (RIM): Review, Architecture, and Experimental Insights
DataFunSummit
DataFunSummit
Jun 14, 2022 · Artificial Intelligence

Practical Acceleration of Deep Model Inference: Case Studies and Optimization Techniques

This talk presents practical methods for accelerating deep model inference, detailing two case studies—text QA and speech QA—along with their technical challenges, and outlines optimization strategies such as model compression, multi‑operator fusion, matrix multiplication tuning, quantization, and dynamic batching.

Deep LearningDynamic BatchingInference Acceleration
0 likes · 12 min read
Practical Acceleration of Deep Model Inference: Case Studies and Optimization Techniques
DaTaobao Tech
DaTaobao Tech
Jun 14, 2022 · Artificial Intelligence

Six Alibaba CVPR 2022 Papers: Summaries and Open‑Source Releases

Alibaba’s DaTaobao team presented six CVPR 2022 papers—covering video restoration, real‑time image enhancement, human‑object interaction detection, indirect illumination modeling, editable NeRFs, and robust NeRF reconstruction—with all code and models open‑sourced, and invites viewers to a live stream for detailed walkthroughs.

CVPRDeep LearningNeRF
0 likes · 5 min read
Six Alibaba CVPR 2022 Papers: Summaries and Open‑Source Releases
Xiaohongshu Tech REDtech
Xiaohongshu Tech REDtech
Jun 13, 2022 · Artificial Intelligence

Neighbor Transformer (NFormer): Robust Person Re-identification via Interactive Multi‑image Modeling

Neighbor Transformer (NFormer) introduces interactive multi‑image modeling for person re‑identification, using Landmark Agent Attention and Reciprocal Neighbor Softmax to efficiently fuse features across images, achieving state‑of‑the‑art accuracy and tighter embedding clusters on multiple benchmark datasets.

Computer VisionDeep Learninglandmark agent attention
0 likes · 8 min read
Neighbor Transformer (NFormer): Robust Person Re-identification via Interactive Multi‑image Modeling
ITPUB
ITPUB
Jun 9, 2022 · Artificial Intelligence

How 58’s Multi‑Label Image Recognition Boosts Semantic Search and Recommendations

This article details the design, data pipeline, model architecture, loss functions, and evaluation metrics of a large‑scale multi‑label image classification system built for 58.com, showing how it improves semantic similarity detection, recommendation, and content moderation across diverse business domains.

Computer VisionDeep Learningasymmetric loss
0 likes · 18 min read
How 58’s Multi‑Label Image Recognition Boosts Semantic Search and Recommendations
Code DAO
Code DAO
May 31, 2022 · Artificial Intelligence

How Deep Convolutional Networks Boost Image Super-Resolution: A Paper Review

This article reviews the seminal SRCNN paper, detailing its contributions, architecture, training pipeline, hyper‑parameters, and extensive experiments that show how a shallow fully‑convolutional network achieves superior PSNR and runtime compared to traditional sparse‑coding and bicubic methods.

CNNDeep LearningPSNR
0 likes · 12 min read
How Deep Convolutional Networks Boost Image Super-Resolution: A Paper Review
DataFunTalk
DataFunTalk
May 28, 2022 · Artificial Intelligence

Adversarial Examples for Captcha: Techniques, Applications, and Future Directions

This article presents a comprehensive overview of adversarial example research applied to captcha systems, covering the definition and history of adversarial attacks, geometric‑aware generation frameworks, FGSM‑based attack variants, experimental results, trade‑offs between image quality and attack strength, and future work such as AdvGAN integration.

AI SafetyDeep LearningFGSM
0 likes · 14 min read
Adversarial Examples for Captcha: Techniques, Applications, and Future Directions
NetEase LeiHuo Testing Center
NetEase LeiHuo Testing Center
May 27, 2022 · Artificial Intelligence

Multimodal Model for Game Frame Rate Prediction

This article explains how a multimodal deep learning model combines static and temporal game data to predict frame rates, helping identify performance bottlenecks and improve client smoothness through feature fusion, data pipelines, and real‑time inference in modern games.

AIDeep LearningMultimodal Learning
0 likes · 7 min read
Multimodal Model for Game Frame Rate Prediction
Hulu Beijing
Hulu Beijing
May 26, 2022 · Artificial Intelligence

Why Vector Retrieval Outperforms Keyword Search for Personalized Video Discovery

This article explains how modern video platforms combine traditional keyword retrieval with deep‑learning‑based vector retrieval, detailing model architectures, attention mechanisms, personalization features, offline experiments, and online A/B results that show significant improvements in recall, relevance, and user experience.

Deep LearningVector Retrievalinformation retrieval
0 likes · 18 min read
Why Vector Retrieval Outperforms Keyword Search for Personalized Video Discovery
AntTech
AntTech
May 24, 2022 · Artificial Intelligence

WPipe: Group‑Based Interleaved Pipeline Parallelism for Large‑Scale DNN Training

The paper introduces WPipe, a group‑based interleaved pipeline parallelism method that reduces memory overhead and weight‑update latency compared with PipeDream‑2BW, achieving up to 1.4× speed‑up and 36% lower memory usage while preserving model accuracy on large‑scale DNNs.

Deep LearningPipeline ParallelismTraining Throughput
0 likes · 13 min read
WPipe: Group‑Based Interleaved Pipeline Parallelism for Large‑Scale DNN Training
DataFunTalk
DataFunTalk
May 23, 2022 · Artificial Intelligence

A Survey of Deep Matching Models for Search and Recommendation

This article surveys recent deep learning approaches for matching in search and recommendation systems, presenting a unified view of matching, categorizing methods into representation learning and matching function learning, and detailing model architectures from input to output layers, while highlighting broader applications such as QA and image captioning.

Deep LearningSearchmatching
0 likes · 4 min read
A Survey of Deep Matching Models for Search and Recommendation
DataFunTalk
DataFunTalk
May 14, 2022 · Artificial Intelligence

Introducing DGL: An Efficient, User‑Friendly, Open Graph Deep Learning Platform

This article presents an overview of graph data and graph neural networks, explains the core concepts of message‑passing GNNs, highlights DGL’s flexible API, high‑performance system design, large‑scale training capabilities and open‑source ecosystem, and outlines future plans and community resources.

DGLDeep Learninggraph data
0 likes · 17 min read
Introducing DGL: An Efficient, User‑Friendly, Open Graph Deep Learning Platform
DataFunTalk
DataFunTalk
May 14, 2022 · Artificial Intelligence

Call for Papers: 4th International Workshop on Deep Learning Practice for High‑Dimensional Sparse Data (DLP‑KDD 2022)

The 4th International Workshop on Deep Learning Practice for High‑Dimensional Sparse and Imbalanced Data (DLP‑KDD 2022) invites submissions on deep‑learning systems, data representation, and user modeling for large‑scale sparse data, with a deadline of May 26, 2022 and acceptance notifications by June 20, 2022.

AIDeep LearningSparse Data
0 likes · 5 min read
Call for Papers: 4th International Workshop on Deep Learning Practice for High‑Dimensional Sparse Data (DLP‑KDD 2022)
Code DAO
Code DAO
May 12, 2022 · Artificial Intelligence

How Activation Functions Work in Deep Learning

This article explains the role of activation functions in deep learning, covering their definition, why they are needed, the main categories—including linear, binary step, and various non‑linear functions such as Sigmoid, TanH, ReLU, Leaky ReLU, ELU, Softmax and Swish—along with each function's mathematical form, advantages, disadvantages, and practical usage recommendations.

Deep LearningNeural NetworkReLU
0 likes · 13 min read
How Activation Functions Work in Deep Learning
Baidu Geek Talk
Baidu Geek Talk
May 6, 2022 · Artificial Intelligence

Artificial Intelligence Development History and Pre‑training Model Trends

From the 1940s birth of computers to today's ultra‑large pre‑training models like Baidu’s ERNIE 3.0, AI has progressed through three development waves, now driven by algorithms, compute and data, with pre‑training lowering application barriers and evolving toward larger, multimodal, and more generalizable systems.

Artificial IntelligenceDeep Learningmachine learning
0 likes · 11 min read
Artificial Intelligence Development History and Pre‑training Model Trends
DataFunTalk
DataFunTalk
May 4, 2022 · Artificial Intelligence

Advances in Recommendation Models: CTR Prediction, Continuous Feature Embedding, Interaction Modeling, and Distributed Training

This article reviews the evolution of recommendation models from early collaborative filtering to modern deep learning approaches, discusses core challenges such as CTR prediction, outlines user‑behavior and combination‑feature modeling techniques, introduces large‑embedding training and continuous‑feature embedding methods like AutoDis, and presents distributed training frameworks such as ScaleFreeCTR, concluding with future research directions.

CTR predictionDeep LearningEmbedding
0 likes · 21 min read
Advances in Recommendation Models: CTR Prediction, Continuous Feature Embedding, Interaction Modeling, and Distributed Training
Tencent Cloud Developer
Tencent Cloud Developer
Apr 27, 2022 · Artificial Intelligence

Alignment-Uniformity Representation Learning for Zero-shot Video Classification (AURL)

The AURL framework, presented by Pu Shi, introduces alignment‑uniformity aware representation learning for zero‑shot video classification, achieving up to 28 % top‑1 accuracy gains on UCF101 and HMDB51, and has already boosted business metrics in Tencent’s advertising, search, and video‑channel recommendation systems.

AlignmentComputer VisionDeep Learning
0 likes · 19 min read
Alignment-Uniformity Representation Learning for Zero-shot Video Classification (AURL)
Alibaba Cloud Developer
Alibaba Cloud Developer
Apr 26, 2022 · Artificial Intelligence

Unlocking Vision AI: Inside Alibaba’s EasyCV All‑in‑One Self‑Supervised & Transformer Framework

EasyCV is Alibaba’s open‑source, PyTorch‑based visual modeling platform that unifies self‑supervised learning and Transformer techniques, offering a comprehensive algorithm suite, pre‑trained models, high‑performance training/inference optimizations, extensible architecture, and seamless cloud deployment for a wide range of computer‑vision tasks.

AI FrameworkAlibabaDeep Learning
0 likes · 16 min read
Unlocking Vision AI: Inside Alibaba’s EasyCV All‑in‑One Self‑Supervised & Transformer Framework
Code DAO
Code DAO
Apr 24, 2022 · Artificial Intelligence

How Transfer Learning Accelerates Deep Learning Across Vision, NLP, and Reinforcement Learning

The article explains how transfer learning reduces data and time requirements in deep learning by reusing pretrained models for vision, natural language processing, and reinforcement learning, while discussing challenges such as overfitting, the need for progressive networks, entropy regularization, domain adaptation, multi‑task learning, and model distillation.

Deep LearningReinforcement Learningdomain adaptation
0 likes · 10 min read
How Transfer Learning Accelerates Deep Learning Across Vision, NLP, and Reinforcement Learning
360 Quality & Efficiency
360 Quality & Efficiency
Apr 22, 2022 · Artificial Intelligence

Audio Quality Assessment Using a BiLSTM Deep Learning Model

This article presents a no‑reference audio quality assessment system that leverages a bidirectional LSTM network to extract spectral features via FFT and predict perceptual scores, describing the architecture, technical advantages, data preparation, loss design, and TensorFlow implementation.

BiLSTMDeep LearningSignal Processing
0 likes · 8 min read
Audio Quality Assessment Using a BiLSTM Deep Learning Model
Programmer DD
Programmer DD
Apr 18, 2022 · Artificial Intelligence

Unlocking Captcha Secrets: How the Open‑Source ddddocr Python Library Works

This article introduces the open‑source Python library ddddocr, explains its evolution from version 1.2.0 to 1.4.3—including OCR, target detection, and slider recognition features—and shows how it leverages deep‑learning and OpenCV to simplify captcha solving for developers.

CaptchaDeep LearningOCR
0 likes · 4 min read
Unlocking Captcha Secrets: How the Open‑Source ddddocr Python Library Works
GuanYuan Data Tech Team
GuanYuan Data Tech Team
Apr 14, 2022 · Artificial Intelligence

Mastering Time Series Forecasting: From Moving Averages to Transformers

Time series forecasting, essential across weather, finance, and commerce, involves tasks like classification, clustering, anomaly detection, and especially prediction; this article explores its definitions, evaluation metrics, traditional methods, machine‑learning approaches, deep‑learning models such as TFT, and emerging AutoML tools, offering practical insights and best practices.

AutoMLDeep LearningGBDT
0 likes · 27 min read
Mastering Time Series Forecasting: From Moving Averages to Transformers
DataFunSummit
DataFunSummit
Apr 12, 2022 · Artificial Intelligence

Intelligent Auction Mechanisms for Alibaba Display Advertising: AIDA Framework, Deep GSP, and Neural Auction

This article presents the evolution of Alibaba's display advertising auction mechanisms, introducing the AIDA decision‑allocation framework, the Deep GSP multi‑objective smart auction, and the end‑to‑end Neural Auction, while discussing their economic theory, engineering platformization, business impact, and future research directions.

AIDeep Learningauction
0 likes · 18 min read
Intelligent Auction Mechanisms for Alibaba Display Advertising: AIDA Framework, Deep GSP, and Neural Auction
Kuaishou Tech
Kuaishou Tech
Apr 11, 2022 · Artificial Intelligence

Kuaishou's Custom Video Matting Solution: Interactive Object Segmentation for Mobile Creators

Kuaishou's audio‑video technology team presents a self‑developed custom video matting system that combines foreground, interactive, and video object segmentation to let creators extract arbitrary subjects without green screens, featuring adaptive cropping, multi‑stage training, and deployment across Android and iOS devices.

Computer VisionDeep LearningKuaishou
0 likes · 15 min read
Kuaishou's Custom Video Matting Solution: Interactive Object Segmentation for Mobile Creators
Baidu Geek Talk
Baidu Geek Talk
Apr 8, 2022 · Artificial Intelligence

Golang Object Pool for Reducing GC Pressure, FFmpeg Concurrency Control, and Paddle Static vs. Dynamic Graphs

The article explains how Go's lock‑free sync.Pool can cut garbage‑collection overhead, shows practical FFmpeg thread‑parameter tuning that balances CPU use and latency for video filtering versus encoding, and compares PaddlePaddle's static and dynamic graph modes, including debugging tips and conversion to static.

Deep LearningDynamic GraphStatic Graph
0 likes · 13 min read
Golang Object Pool for Reducing GC Pressure, FFmpeg Concurrency Control, and Paddle Static vs. Dynamic Graphs
Kuaishou Large Model
Kuaishou Large Model
Apr 6, 2022 · Artificial Intelligence

How Transformers Revolutionize Image Style Transfer: Introducing StyTr²

This article reviews the limitations of traditional CNN‑based image stylization, explains how Transformer architectures overcome these issues with global context and self‑attention, and presents the novel StyTr² method with content‑aware positional encoding that achieves superior, detail‑preserving style transfer results.

Computer VisionDeep LearningTransformer
0 likes · 8 min read
How Transformers Revolutionize Image Style Transfer: Introducing StyTr²
Kuaishou Tech
Kuaishou Tech
Apr 6, 2022 · Artificial Intelligence

StyTr²: A Transformer‑Based Approach for Image Style Transfer

The paper proposes StyTr², a Transformer‑based image style transfer method that uses content‑aware positional encoding to preserve details and improve feature representation, achieving high‑quality stylization with better content structure and style patterns.

Computer VisionDeep Learningcontent-aware positional encoding
0 likes · 7 min read
StyTr²: A Transformer‑Based Approach for Image Style Transfer
DataFunTalk
DataFunTalk
Apr 6, 2022 · Artificial Intelligence

AIDA Advertising Intelligent Decision and Allocation Framework: Evolution of Smart Auction Mechanisms

This article introduces the AIDA framework for Alibaba's display advertising, detailing the business background, multi‑objective optimization challenges, the design of Deep GSP and Neural Auction mechanisms powered by deep learning and reinforcement learning, and outlines future technical and platform directions while also announcing recruitment opportunities.

AIAdvertisingDeep Learning
0 likes · 16 min read
AIDA Advertising Intelligent Decision and Allocation Framework: Evolution of Smart Auction Mechanisms
NetEase LeiHuo Testing Center
NetEase LeiHuo Testing Center
Apr 1, 2022 · Artificial Intelligence

Learning OCR for Game Text Recognition: From Data Preparation to CRNN Model Training

This article documents the author’s step‑by‑step journey of building an OCR system for recognizing Chinese characters in a card‑game UI, covering game selection, technical background, data generation, deep‑learning model training with CRNN, real‑image data collection, optimization attempts, and final performance evaluation.

CRNNDeep LearningEasyOCR
0 likes · 15 min read
Learning OCR for Game Text Recognition: From Data Preparation to CRNN Model Training
Python Programming Learning Circle
Python Programming Learning Circle
Mar 31, 2022 · Artificial Intelligence

Comprehensive PyTorch Code Snippets: Configuration, Tensor Operations, Model Definition, Training, and Best Practices

This article provides a thorough collection of commonly used PyTorch code snippets covering environment setup, reproducibility, GPU configuration, tensor manipulation, model building, data preprocessing, training and evaluation loops, custom loss functions, regularization techniques, learning‑rate scheduling, checkpointing, and practical tips for efficient deep‑learning development.

Deep LearningGPUModel Training
0 likes · 37 min read
Comprehensive PyTorch Code Snippets: Configuration, Tensor Operations, Model Definition, Training, and Best Practices
Cyber Elephant Tech Team
Cyber Elephant Tech Team
Mar 30, 2022 · Artificial Intelligence

Can AI Make Real-Life Invisibility Cloaks? Inside the STTN Video Restoration Breakthrough

This article reviews the challenges of video inpainting, surveys traditional methods, and introduces the Spatial‑Temporal Transformer Network (STTN) that leverages multi‑scale attention and a Temporal Patch‑GAN discriminator, detailing its architecture, loss functions, training on Youtube‑VOS, and impressive restoration results.

AI video restorationDeep LearningVideo Inpainting
0 likes · 10 min read
Can AI Make Real-Life Invisibility Cloaks? Inside the STTN Video Restoration Breakthrough
Baidu Geek Talk
Baidu Geek Talk
Mar 28, 2022 · Artificial Intelligence

Robust Input Visualization Methods for Vision Transformers

The paper proposes a robust Grad‑CAM‑inspired visualization for Vision Transformers that combines attention weights and gradients to generate class‑specific saliency maps, demonstrates superior alignment with discriminative regions across ViT, Swin and Volo models, and shows a 76% false‑positive reduction in Baidu’s porn‑content risk control system.

Deep LearningGrad-CAMInput Visualization
0 likes · 11 min read
Robust Input Visualization Methods for Vision Transformers
DataFunSummit
DataFunSummit
Mar 26, 2022 · Artificial Intelligence

Deep Learning‑Based Design of Financial Index Funds Using Graph Neural Networks

This talk presents a deep‑learning framework that formulates financial index‑fund construction as a sparse portfolio optimization problem, solves the mixed‑integer programming via a two‑stage graph‑neural‑network pipeline, and demonstrates superior tracking performance and scalability on large‑scale index datasets.

AI FinanceDeep Learningfinancial index funds
0 likes · 16 min read
Deep Learning‑Based Design of Financial Index Funds Using Graph Neural Networks
Laiye Technology Team
Laiye Technology Team
Mar 25, 2022 · Artificial Intelligence

Laiye OCR Error‑Correction Model: Architecture, Implementation, and Evaluation

This article describes Laiye's OCR error‑correction system, detailing the background challenges of Chinese character recognition, the analysis of three possible solutions, the chosen post‑processing approach, model architecture, training data, loss design, online inference, and experimental results showing a measurable performance boost.

Chinese textComputer VisionDeep Learning
0 likes · 13 min read
Laiye OCR Error‑Correction Model: Architecture, Implementation, and Evaluation
Meituan Technology Team
Meituan Technology Team
Mar 24, 2022 · Artificial Intelligence

Cyclic Generative Adversarial Networks for Probability Density Estimation – Academic Salon by Tsinghua University & Meituan Digital Life

The Tsinghua‑Meituan Digital Life Joint Research Institute’s academic salon will feature Associate Professor Jiang Rui presenting a cyclic generative adversarial network for probability density estimation, demonstrating how merging statistical models with deep‑learning techniques can solve core statistical problems and foster industry‑academia innovation.

Artificial IntelligenceDeep LearningGenerative Adversarial Networks
0 likes · 4 min read
Cyclic Generative Adversarial Networks for Probability Density Estimation – Academic Salon by Tsinghua University & Meituan Digital Life
JD Cloud Developers
JD Cloud Developers
Mar 21, 2022 · Artificial Intelligence

ViTAEv2 Breaks ImageNet Real Record with 91.2% Accuracy – How a 600M‑Parameter Model Redefines Few‑Shot Learning

JD Research Institute and the University of Sydney introduced ViTAEv2, a 600‑million‑parameter deep learning model that achieved a world‑leading 91.2% top‑1 accuracy on ImageNet Real without external data, demonstrating strong few‑shot learning, reducing labeling costs, and promising advances across many computer‑vision tasks.

AI modelComputer VisionDeep Learning
0 likes · 4 min read
ViTAEv2 Breaks ImageNet Real Record with 91.2% Accuracy – How a 600M‑Parameter Model Redefines Few‑Shot Learning
DataFunSummit
DataFunSummit
Mar 16, 2022 · Artificial Intelligence

Semantic Search Recall Techniques at JD: Dual‑Tower Model, Graph Model, Synonym Recall, and Joint Index Training

This article presents JD's end‑to‑end semantic search recall pipeline, covering multi‑stage recall, a dual‑tower embedding model with multi‑head attention, a heterogeneous graph neural network (SearchGCN), a transformer‑based synonym generation system, and a joint index‑training approach that integrates product quantization to improve recall accuracy and efficiency.

Deep LearningGraph Neural Networkdual-tower model
0 likes · 17 min read
Semantic Search Recall Techniques at JD: Dual‑Tower Model, Graph Model, Synonym Recall, and Joint Index Training
DataFunTalk
DataFunTalk
Mar 16, 2022 · Artificial Intelligence

Parameter-Efficient Sparsity Training for the PLUG Large-Scale Language Model

This article presents the PLUG 270‑billion‑parameter Chinese language model and introduces a parameter‑efficient sparsity training (PST) framework that combines unstructured and structured pruning with low‑rank decomposition to dramatically reduce model size while preserving downstream performance.

Deep LearningLarge Language ModelsPLUG
0 likes · 13 min read
Parameter-Efficient Sparsity Training for the PLUG Large-Scale Language Model
Tencent Cloud Developer
Tencent Cloud Developer
Mar 15, 2022 · Artificial Intelligence

Comprehensive Overview of Ranking Models in Recommendation Systems

The article provides a thorough guide to ranking in recommendation systems, detailing the pipeline architecture, sample handling challenges, extensive feature engineering categories, the evolution from collaborative filtering to deep and attention‑based models, and key optimization trade‑offs between memorization, generalization, and efficient user‑interest modeling.

CTR predictionDeep LearningModel Optimization
0 likes · 19 min read
Comprehensive Overview of Ranking Models in Recommendation Systems
DataFunTalk
DataFunTalk
Mar 12, 2022 · Artificial Intelligence

NetEase Cloud Music Advertising System: Algorithm Practice and Model Evolution

This article presents a comprehensive overview of NetEase Cloud Music's advertising system, detailing its architecture, core challenges, CTR and CVR prediction models, feature engineering, model evolution from LR to deep learning, user vector modeling, and practical recommendations for improving ad performance.

AdvertisingCTR predictionDeep Learning
0 likes · 15 min read
NetEase Cloud Music Advertising System: Algorithm Practice and Model Evolution
DeWu Technology
DeWu Technology
Mar 11, 2022 · Artificial Intelligence

Deep Learning in Face Recognition

The article surveys deep‑learning‑based face‑recognition systems, detailing detection, preprocessing, and recognition pipelines, describing evaluation metrics such as TAR, FAR, and Rank‑K, reviewing major datasets like LFW, MS‑Celeb‑1M and VGGFace2, and comparing leading architectures—including FaceNet, CenterLoss, SphereFace and InsightFace—while highlighting their strengths, limitations, real‑world applications, and seminal research references.

AIDatasetsDeep Learning
0 likes · 14 min read
Deep Learning in Face Recognition
DaTaobao Tech
DaTaobao Tech
Mar 11, 2022 · Artificial Intelligence

How Alibaba’s MNN Engine Achieves 350% CPU Speedup and Sparse Acceleration

Alibaba’s MNN, a lightweight high‑performance deep‑learning inference engine, earned top honors in China’s 2022 “Science & Innovation China” awards, and delivers impressive gains such as 350% speedup on X86 CPUs, 2.1‑2.3× acceleration on ARM with sparse models, plus integrated OpenCV/Numpy functionality for edge AI deployment.

AI deploymentAlibabaDeep Learning
0 likes · 4 min read
How Alibaba’s MNN Engine Achieves 350% CPU Speedup and Sparse Acceleration
DataFunTalk
DataFunTalk
Mar 9, 2022 · Artificial Intelligence

Semantic Search Recall Techniques at JD: Dual‑Tower Model, Graph Model, Synonym Recall, and Index Joint Training

The talk presents JD's end‑to‑end semantic search recall pipeline, covering multi‑stage retrieval, a dual‑tower embedding model with multi‑head attention, a heterogeneous graph neural network for low‑frequency items, automatic synonym generation via transformer models, and a joint training approach that integrates product quantization directly into the model to improve accuracy and efficiency.

Deep LearningGraph Neural Networkdual-tower model
0 likes · 16 min read
Semantic Search Recall Techniques at JD: Dual‑Tower Model, Graph Model, Synonym Recall, and Index Joint Training
DataFunSummit
DataFunSummit
Mar 6, 2022 · Artificial Intelligence

The Evolution of Embedding Techniques: From Word2Vec to Graph Neural Networks

This article traces the development of embedding methods—from the early word2vec model through item2vec, DeepWalk, Node2vec, EGES, HERec, GraphRT, and target‑fitting approaches like DSSM and YouTube recommendation—highlighting how sequence‑construction and target‑fitting paradigms have shaped modern recommendation systems and AI applications.

Deep LearningEmbeddingItem2Vec
0 likes · 26 min read
The Evolution of Embedding Techniques: From Word2Vec to Graph Neural Networks
Alibaba Cloud Native
Alibaba Cloud Native
Mar 5, 2022 · Cloud Native

How Fluid Accelerates Cloud‑Native Deep Learning Training

Fluid, an open‑source CNCF project co‑developed by Alibaba Cloud and Nanjing University, introduces a dataset abstraction and elastic caching architecture that automatically optimizes I/O for cloud‑native deep‑learning training jobs, and its research was accepted as a full paper at the prestigious ICDE 2022 conference.

Cloud NativeData AccelerationDeep Learning
0 likes · 6 min read
How Fluid Accelerates Cloud‑Native Deep Learning Training
IT Services Circle
IT Services Circle
Mar 2, 2022 · Artificial Intelligence

Curated Open‑Source Face Recognition Projects Overview

This article presents a curated collection of open‑source face recognition projects—including OpenFace, face_recognition, InsightFace, RetinaFace, SCRFD, FaceNet, Deepface, and CompreFace—detailing their features, GitHub stars, usage examples, and code snippets for Python and TensorFlow implementations.

Artificial IntelligenceDeep LearningPython
0 likes · 5 min read
Curated Open‑Source Face Recognition Projects Overview
DataFunSummit
DataFunSummit
Mar 1, 2022 · Artificial Intelligence

Alibaba's Smart Supply‑Chain Forecasting: Scenarios, Algorithm R&D, and Application Cases

This article details Alibaba's exploration of intelligent supply‑chain forecasting, covering scenario classification, three generations of prediction algorithms, the self‑developed Falcon model, performance evaluation, and real‑world cases such as Double 11 and live‑streaming, highlighting challenges and practical solutions.

AIDeep LearningTime Series
0 likes · 18 min read
Alibaba's Smart Supply‑Chain Forecasting: Scenarios, Algorithm R&D, and Application Cases
DataFunSummit
DataFunSummit
Feb 26, 2022 · Artificial Intelligence

Graph-Based Sparse Behavior Recall Models for Content Recommendation

This article presents a comprehensive study of graph‑based recall techniques for content recommendation, detailing how knowledge‑graph‑augmented user‑behavior graphs and novel attention‑driven models such as GADM, SGGA, and SGGGA improve performance for users with sparse interaction histories.

Attention MechanismDeep LearningRecommendation Systems
0 likes · 11 min read
Graph-Based Sparse Behavior Recall Models for Content Recommendation
iQIYI Technical Product Team
iQIYI Technical Product Team
Feb 25, 2022 · Artificial Intelligence

Short Video Content Tagging: Multimodal AI Model Framework and Applications

The framework tags short videos by fusing text, image and audio‑video features through specialized extraction, classification, generative and retrieval modules, then ranking candidates with a multimodal BERT model, delivering accurate, business‑specific tags that boost recommendation, search and advertising.

Deep LearningMultimodal AIcontent tagging
0 likes · 10 min read
Short Video Content Tagging: Multimodal AI Model Framework and Applications
Kuaishou Tech
Kuaishou Tech
Feb 18, 2022 · Game Development

Motion Retargeting Techniques for High‑Quality Virtual Character Driving

The article surveys motion retargeting, describing its importance for virtual characters, outlining traditional geometric methods and recent deep‑learning approaches, presenting a customized Interaction Mesh‑based solution from Kuaishou Y‑tech, and discussing performance, limitations, and future research directions.

Deep Learningmotion retargetingvirtual characters
0 likes · 11 min read
Motion Retargeting Techniques for High‑Quality Virtual Character Driving
DataFunTalk
DataFunTalk
Feb 18, 2022 · Artificial Intelligence

Travel Intent Prediction in E-commerce: Algorithm Strategies, Multi‑source Behavior Modeling, and Model Design

This talk presents Alibaba's travel intent prediction system, detailing the unique challenges of low‑frequency, multi‑source travel behavior, the multi‑granular CNN and time‑attention model architecture, experimental comparisons with baselines, and how integrated user interest modeling improves recommendation performance.

Deep Learningattentionmachine learning
0 likes · 11 min read
Travel Intent Prediction in E-commerce: Algorithm Strategies, Multi‑source Behavior Modeling, and Model Design
Baidu Geek Talk
Baidu Geek Talk
Feb 17, 2022 · Artificial Intelligence

AI-Powered Sports Video Applications: Figure Skating Action Recognition, Multimodal Classification, and Football Highlight Clipping

The article showcases three AI‑driven sports video solutions—real‑time figure‑skating action recognition with ST‑GCN, multimodal video classification merging text, image and audio via ERNIE and TextCNN, and automated football highlight clipping using TSN‑BMN‑LSTM—each achieving over 85% accuracy, fully open‑source on PaddlePaddle with one‑click notebooks and a live developer session.

AIDeep LearningMultimodal Classification
0 likes · 8 min read
AI-Powered Sports Video Applications: Figure Skating Action Recognition, Multimodal Classification, and Football Highlight Clipping
DataFunSummit
DataFunSummit
Feb 14, 2022 · Artificial Intelligence

Evolution of 58 Local Service Recommendation Algorithms and Future Directions

This article presents a comprehensive overview of 58's local service recommendation system, detailing the characteristics of its recommendation scenarios, the evolution of tag and post recommendation pipelines, the underlying deep‑learning models such as Bi‑LSTM, ATRank, DeepFM and ESMM, and outlines future research directions.

ATRankCTRCVR
0 likes · 16 min read
Evolution of 58 Local Service Recommendation Algorithms and Future Directions
Kuaishou Large Model
Kuaishou Large Model
Feb 11, 2022 · Artificial Intelligence

How Motion Retargeting Powers Real‑Time Virtual Humans: Techniques & Insights

Motion retargeting, also known as motion adaptation, transfers source motion to arbitrary virtual characters while preserving semantic features and natural flow, and this article reviews traditional geometry‑based methods, recent deep‑learning approaches, and Kuaishou Y‑Tech’s optimized pipeline that balances quality, contact preservation, and real‑time performance.

Deep Learninganimationcomputer graphics
0 likes · 13 min read
How Motion Retargeting Powers Real‑Time Virtual Humans: Techniques & Insights
DataFunTalk
DataFunTalk
Feb 5, 2022 · Artificial Intelligence

Evolution of 58 Local Service Recommendation Algorithms: Scenarios, Tag & Post Recommendations, and Future Directions

This article presents a comprehensive overview of 58 Local Service's recommendation system, detailing the diverse recommendation scenarios, challenges such as information homogeneity and complex user structures, the multi‑stage recall and ranking pipelines, model evolutions from statistical methods to deep learning, and future work to improve data quality and model efficiency.

ATRankCTRCVR
0 likes · 15 min read
Evolution of 58 Local Service Recommendation Algorithms: Scenarios, Tag & Post Recommendations, and Future Directions
DataFunTalk
DataFunTalk
Feb 2, 2022 · Artificial Intelligence

UGC Sentiment Analysis Solutions and Applications in Taobao

This article presents a comprehensive overview of Taobao's user‑generated content sentiment analysis pipeline, covering task definition, challenges, model architecture with RoBERTa‑based extraction, sentiment‑knowledge pre‑training, graph augmentation, personalized ranking, business impact metrics, and future research directions.

Deep LearningSentiment AnalysisUGC
0 likes · 16 min read
UGC Sentiment Analysis Solutions and Applications in Taobao
MaGe Linux Operations
MaGe Linux Operations
Jan 30, 2022 · Artificial Intelligence

PyTorch vs TensorFlow in 2022: Which Framework Wins for Your Needs?

This article compares PyTorch and TensorFlow in 2022 across model availability, deployment ease, and ecosystem support, using data from HuggingFace, research papers, and industry tools, and offers tailored recommendations for industry engineers, researchers, educators, career changers, hobbyists, and beginners.

AIDeep LearningModel Deployment
0 likes · 20 min read
PyTorch vs TensorFlow in 2022: Which Framework Wins for Your Needs?
DataFunSummit
DataFunSummit
Jan 29, 2022 · Artificial Intelligence

Survey of Model Pruning and Quantization Techniques for Deep Learning

This article provides a comprehensive overview of recent advances in deep learning model compression, focusing on pruning methods—including unstructured, structured, filter-wise, channel-wise, shape-wise, and stripe-wise approaches—and quantization techniques such as linear, non‑linear, clustering, power‑of‑two, binary, and 8‑bit quantization, while discussing evaluation criteria, sparsity ratios, fine‑tuning, and training‑aware quantization.

Deep LearningNeural Networksmodel compression
0 likes · 23 min read
Survey of Model Pruning and Quantization Techniques for Deep Learning
Xianyu Technology
Xianyu Technology
Jan 29, 2022 · Artificial Intelligence

Semantic Vector Retrieval and I2I Recall Optimization in Xianyu Search

Xianyu search recall stage upgraded from simple text matching to semantic vector retrieval using DSSM with Electra‑Small, query graph attention, and behavior‑based I2I, adding structured attributes and OCR tags, improving AUC to 0.824 and HitRate@10 to 90.1%, boosting purchase metrics by up to 4%.

Deep LearningVector RetrievalXianyu
0 likes · 17 min read
Semantic Vector Retrieval and I2I Recall Optimization in Xianyu Search
Bilibili Tech
Bilibili Tech
Jan 28, 2022 · Artificial Intelligence

Real-CUGAN: An Open‑Source AI Super‑Resolution Model for Anime Video Upscaling

Real‑CUGAN is an open‑source AI super‑resolution model that upscales anime video up to 4× using a million‑patch, frequency‑domain‑supervised dataset, delivering faster inference than Real‑ESRGAN, seamless Waifu2x compatibility, and superior texture, line and artifact handling, with code released on GitHub.

AI super-resolutionDeep LearningImage Restoration
0 likes · 8 min read
Real-CUGAN: An Open‑Source AI Super‑Resolution Model for Anime Video Upscaling
Laiye Technology Team
Laiye Technology Team
Jan 28, 2022 · Artificial Intelligence

Survey of Model Compression and Quantization Techniques for Deep Neural Networks

This article provides a comprehensive overview of deep learning model compression and acceleration methods, detailing pruning strategies, various pruning types, evaluation criteria, sparsity ratios, fine‑tuning procedures, as well as linear and non‑linear quantization approaches, their implementations, and practical considerations.

Deep LearningNeural Networksefficiency
0 likes · 26 min read
Survey of Model Compression and Quantization Techniques for Deep Neural Networks
Baobao Algorithm Notes
Baobao Algorithm Notes
Jan 28, 2022 · Artificial Intelligence

How Pre‑Training Evolved: From word2vec to MAE Across NLP and CV

This article traces the history of deep‑learning pre‑training techniques, comparing the parallel developments in natural‑language processing and computer vision—from early word2vec and bag‑of‑words models through ELMo and BERT to recent transformer‑based vision models like iGPT, ViT, BEiT and MAE—highlighting key innovations, challenges, and the convergence of the two fields.

Deep LearningMAENLP
0 likes · 20 min read
How Pre‑Training Evolved: From word2vec to MAE Across NLP and CV
DataFunTalk
DataFunTalk
Jan 23, 2022 · Artificial Intelligence

Dual-Sequence Fusion for New‑User Cold‑Start Recall in Content Recommendation

This article presents a systematic study of recall techniques for new‑user cold‑start in content recommendation, describing a baseline two‑tower model, a Dual Attention Network (DAN) fusion approach, and an enhanced Contextual‑Gate DAN that dynamically balances content and product sequences, together with offline and online evaluation results and future directions.

Deep LearningUser Embeddingcold-start
0 likes · 12 min read
Dual-Sequence Fusion for New‑User Cold‑Start Recall in Content Recommendation
Python Programming Learning Circle
Python Programming Learning Circle
Jan 18, 2022 · Artificial Intelligence

Fashion MNIST Image Classification Using TensorFlow 2.x in Python

This tutorial demonstrates how to load the Fashion MNIST dataset, explore and preprocess the images, build and compile a neural network with TensorFlow 2.x, train the model, evaluate its accuracy, and use the trained model to make predictions on clothing images, providing complete Python code examples throughout.

Deep LearningFashion-MNISTImage Classification
0 likes · 16 min read
Fashion MNIST Image Classification Using TensorFlow 2.x in Python
DataFunSummit
DataFunSummit
Jan 16, 2022 · Artificial Intelligence

Multimodal Text and Speech Emotion Analysis: Overview, MSCNN‑SPU Model, and Domain Adaptation

This talk presents an overview of text‑plus‑speech multimodal emotion analysis, covering background, single‑modal text and audio models, the MSCNN‑SPU multimodal architecture, domain‑adaptation techniques, and future directions, with detailed model explanations, experimental results, and practical deployment insights.

Audio ProcessingDeep Learningmultimodal emotion analysis
0 likes · 40 min read
Multimodal Text and Speech Emotion Analysis: Overview, MSCNN‑SPU Model, and Domain Adaptation
Code DAO
Code DAO
Jan 15, 2022 · Artificial Intelligence

How Intel BF16 with IPEX and oneDNN Boosts PyTorch Performance

This article explains how Intel and Facebook's BF16 support, combined with the Intel Extension for PyTorch (IPEX) and oneDNN, automates type and layout conversions and adds graph‑fusion optimizations, delivering 1.4×‑4.3× inference and up to 2.4× training speedups on Xeon CPUs for models such as DLRM, BERT‑Large, and ResNext‑101‑32x4d.

BF16CPU accelerationDeep Learning
0 likes · 13 min read
How Intel BF16 with IPEX and oneDNN Boosts PyTorch Performance
Baobao Algorithm Notes
Baobao Algorithm Notes
Jan 14, 2022 · Artificial Intelligence

Boosting BERT Text Classification with Label Embedding: How It Works

The paper proposes a simple yet effective method that fuses label embeddings into BERT, enhancing text‑classification performance without increasing computational cost, and validates the approach across six benchmark datasets, also exploring tf‑idf‑based label augmentation and the impact of using [SEP] versus no‑[SEP] inputs.

BERTDeep LearningNLP
0 likes · 8 min read
Boosting BERT Text Classification with Label Embedding: How It Works
Python Programming Learning Circle
Python Programming Learning Circle
Jan 11, 2022 · Artificial Intelligence

Dynamic Learning Rate Adjustment in PyTorch: Optimizer Basics and Scheduler Usage

This article explains how to configure and use PyTorch optimizers, their attributes and methods, and demonstrates various learning‑rate scheduling techniques—including manual updates and built‑in schedulers such as LambdaLR, StepLR, MultiStepLR, ExponentialLR, CosineAnnealingLR, and ReduceLROnPlateau—through clear code examples.

Deep LearningPyTorchScheduler
0 likes · 14 min read
Dynamic Learning Rate Adjustment in PyTorch: Optimizer Basics and Scheduler Usage
DataFunSummit
DataFunSummit
Jan 10, 2022 · Artificial Intelligence

Understanding Vector Retrieval: Principles, Applications, and High‑Performance Algorithms

This article explains how deep learning transforms raw physical‑world data into dense vectors, defines the significance of vector retrieval, surveys common use cases such as image, video, and text search, discusses challenges in representation learning, and reviews high‑performance approximate nearest‑neighbor algorithms and practical deployments.

AI applicationsDeep Learningapproximate nearest neighbor
0 likes · 21 min read
Understanding Vector Retrieval: Principles, Applications, and High‑Performance Algorithms
DataFunTalk
DataFunTalk
Jan 8, 2022 · Artificial Intelligence

Survey of Classic Recommendation Algorithms: LR, FM, FFM, WDL, DeepFM, DCN, and xDeepFM

This article surveys classic recommendation algorithms—including Logistic Regression, Factorization Machines, Field‑aware FM, Wide & Deep, DeepFM, DCN, and xDeepFM—explaining their principles, feature preprocessing, problem scopes, and industrial applications within personalized recommendation systems.

Deep LearningRecommendation Systemsfactorization machines
0 likes · 12 min read
Survey of Classic Recommendation Algorithms: LR, FM, FFM, WDL, DeepFM, DCN, and xDeepFM
Kuaishou Tech
Kuaishou Tech
Jan 7, 2022 · Artificial Intelligence

Transcoded Video Restoration by Temporal‑Spatial Auxiliary Network – AAAI 2022 Paper Overview

The article summarizes a AAAI 2022 paper by Kuaishou's audio‑video algorithm team and Xidian University that introduces a deep‑learning‑based video compression‑artifact restoration method using temporal‑spatial auxiliary supervision, achieving significant PSNR/SSIM gains on transcoded videos without increasing bitrate.

AAAI 2022Deep Learningcompression artifact removal
0 likes · 6 min read
Transcoded Video Restoration by Temporal‑Spatial Auxiliary Network – AAAI 2022 Paper Overview
Baobao Algorithm Notes
Baobao Algorithm Notes
Jan 7, 2022 · Interview Experience

Essential Transformer Interview Cheat Sheet: 11 Must‑Know Q&A

This concise guide presents eleven frequently asked Transformer interview questions with clear, English explanations covering self‑attention formulas, scaling, alternative designs, LayerNorm vs. BatchNorm, positional embeddings, multi‑head mechanisms, and BPE tokenization, helping candidates deliver solid, theory‑backed answers.

BERTDeep LearningLayerNorm
0 likes · 6 min read
Essential Transformer Interview Cheat Sheet: 11 Must‑Know Q&A
Laiye Technology Team
Laiye Technology Team
Jan 7, 2022 · Artificial Intelligence

Understanding Vector Retrieval: Principles, Applications, and High‑Performance Algorithms

This article explains how deep learning transforms unstructured data into dense vectors, defines vector retrieval, outlines its many use cases such as product, video, and text search, discusses challenges in learning effective embeddings, and reviews high‑performance algorithms like LSH, neighbor graphs, and product quantization.

AI applicationsDeep LearningHNSW
0 likes · 21 min read
Understanding Vector Retrieval: Principles, Applications, and High‑Performance Algorithms
JD Cloud Developers
JD Cloud Developers
Jan 4, 2022 · Artificial Intelligence

How JD’s Vega v1 Model Dominated GLUE Benchmark, Surpassing Human Performance

JD Explore’s Vega v1 model topped the GLUE benchmark with a 91.3 average score, outperforming Microsoft, Facebook, and Stanford across multiple NLP tasks, including first‑ever human‑level results on sentiment analysis and coreference, showcasing JD’s leading position in deep‑learning research.

AI researchDeep LearningGLUE benchmark
0 likes · 3 min read
How JD’s Vega v1 Model Dominated GLUE Benchmark, Surpassing Human Performance
Code DAO
Code DAO
Dec 31, 2021 · Artificial Intelligence

Why RegNet Is the Most Flexible Architecture for Computer Vision

RegNet introduces a scalable design space defined by quantized linear functions, enabling flexible trade‑offs between accuracy, efficiency, and mobile deployment, and demonstrates superior performance compared with ResNet, EfficientNet, and other mobile‑optimized networks.

Computer VisionDeep LearningDesign Space
0 likes · 7 min read
Why RegNet Is the Most Flexible Architecture for Computer Vision
Laiye Technology Team
Laiye Technology Team
Dec 31, 2021 · Artificial Intelligence

Overview of Table Recognition Techniques and Practical Implementation

This article reviews the challenges of extracting structured table data from images, compares two‑stage and end‑to‑end OCR approaches, evaluates four state‑of‑the‑art table‑recognition models (SPLERGE, CascadeTabNet, TableMASTER, UnetTable), and presents a practical deployment workflow with performance metrics.

AIComputer VisionDeep Learning
0 likes · 14 min read
Overview of Table Recognition Techniques and Practical Implementation
Python Programming Learning Circle
Python Programming Learning Circle
Dec 27, 2021 · Artificial Intelligence

PyTorch vs TensorFlow in 2022: Which Framework to Choose?

An in‑depth 2022 comparison of PyTorch and TensorFlow evaluates model availability, deployment ease, and ecosystem support, showing PyTorch dominates research while TensorFlow excels in deployment, and offers tailored recommendations for industry professionals, researchers, educators, career changers, hobbyists, and beginners.

AIDeep LearningPyTorch
0 likes · 20 min read
PyTorch vs TensorFlow in 2022: Which Framework to Choose?
Code DAO
Code DAO
Dec 24, 2021 · Artificial Intelligence

Understanding Neural Network Predictions with Integrated Gradients

This article introduces the Integrated Gradients (IG) method for explaining deep neural networks, compares it with saliency maps and Shapley‑based approaches, discusses its axiomatic foundations, and provides a step‑by‑step guide to implementing IG using the open‑source TruLens library, including custom baselines and attribution measures.

Attribution MethodsDeep LearningIntegrated Gradients
0 likes · 14 min read
Understanding Neural Network Predictions with Integrated Gradients
Code DAO
Code DAO
Dec 23, 2021 · Artificial Intelligence

Permutation‑Invariant PIUnet Boosts Multi‑Temporal Satellite Image Super‑Resolution

The article explains how satellite images suffer from limited spatial resolution, why the ordering of multi‑temporal frames is irrelevant, and how the PIUnet model introduces permutation‑invariant equivariant layers to achieve state‑of‑the‑art super‑resolution efficiently, winning the AI4EO challenge.

Deep LearningPIUnetSatellite Imagery
0 likes · 6 min read
Permutation‑Invariant PIUnet Boosts Multi‑Temporal Satellite Image Super‑Resolution
Code DAO
Code DAO
Dec 23, 2021 · Artificial Intelligence

Deep Siamese Network for Measuring Similarity of ECG Signals

This article presents an automated neural‑network framework based on a deep Siamese architecture to learn similarity representations between ECG recordings, covering ECG fundamentals, exploratory data analysis, signal preprocessing, model construction with Keras, and demonstrates how the trained network yields similarity scores applicable to broader signal‑matching tasks.

Deep LearningECGKeras
0 likes · 10 min read
Deep Siamese Network for Measuring Similarity of ECG Signals
DataFunTalk
DataFunTalk
Dec 22, 2021 · Artificial Intelligence

Applying Survival Analysis to User Activity Modeling: Concepts, Methods, and the KwaiSurvival Deep‑Learning Framework

This article explains why traditional DAU metrics are insufficient, introduces survival analysis fundamentals and key functions, demonstrates how Kaplan‑Meier curves can characterize user activity, and presents KwaiSurvival—a deep‑learning‑based survival modeling suite with DeepSurv, DeepHit and N‑MTLR models—for practical user‑engagement and churn‑prevention use cases.

Deep LearningKM curveKwaiSurvival
0 likes · 15 min read
Applying Survival Analysis to User Activity Modeling: Concepts, Methods, and the KwaiSurvival Deep‑Learning Framework
JD Cloud Developers
JD Cloud Developers
Dec 21, 2021 · Artificial Intelligence

How JD Cloud’s Mobile Super‑Resolution SDK Boosts Video Quality and Cuts Bandwidth by 30%

JD Cloud’s new mobile super‑resolution SDK leverages deep‑learning ESPCN algorithms with ROI‑based processing to upscale video streams in real time, delivering up to 80% longer playback, 30% lower bandwidth costs, and measurable quality gains demonstrated through PSNR, VMAF, and SSIM metrics.

Bandwidth ReductionDeep LearningESPCN
0 likes · 6 min read
How JD Cloud’s Mobile Super‑Resolution SDK Boosts Video Quality and Cuts Bandwidth by 30%
DataFunTalk
DataFunTalk
Dec 18, 2021 · Artificial Intelligence

Adaptive Mutual Supervision Multi‑Task Graph Neural Network for Fine‑Grained Urban Traffic Demand Prediction

This work proposes an adaptive mutual‑supervision multi‑task graph neural network that captures spatio‑temporal dynamics and heterogeneous group behaviors to predict fine‑grained urban travel demand, demonstrating over 10% performance gains on real‑world Beijing and Shanghai datasets compared with classic baselines.

Deep LearningGraph Neural NetworkTraffic Prediction
0 likes · 24 min read
Adaptive Mutual Supervision Multi‑Task Graph Neural Network for Fine‑Grained Urban Traffic Demand Prediction
ITPUB
ITPUB
Dec 13, 2021 · Artificial Intelligence

How Data Augmentation Boosts Machine Learning When Data Is Scarce

This article explains how data augmentation can alleviate overfitting by artificially expanding limited training sets, outlines common transformation techniques for images, text, and audio, and discusses the method's benefits, practical applications, and inherent limitations for machine‑learning practitioners.

Computer VisionDeep Learningdata augmentation
0 likes · 6 min read
How Data Augmentation Boosts Machine Learning When Data Is Scarce
DataFunTalk
DataFunTalk
Dec 13, 2021 · Artificial Intelligence

Dual Vector Foil (DVF): Decoupled Index and Model for Large‑Scale Retrieval

The article introduces the Dual Vector Foil (DVF) algorithm system, which decouples index construction from model training to enable lightweight, high‑precision large‑scale recall using arbitrary complex models, and details its two‑stage and one‑stage solutions, graph‑based retrieval implementation, performance optimizations, and experimental results.

Deep LearningRecommendation Systemsalgorithm
0 likes · 28 min read
Dual Vector Foil (DVF): Decoupled Index and Model for Large‑Scale Retrieval
Kuaishou Tech
Kuaishou Tech
Dec 13, 2021 · Artificial Intelligence

AI-Powered High-Resolution Portrait Restoration Using StyleGAN and Face Parsing

This article describes an AI-driven portrait enhancement system that restores degraded facial images by simulating degradation, constructing paired datasets, and employing a StyleGAN‑based generator combined with face‑parsing masks, detailing the pipeline, model architecture, training losses, and achieved high‑quality results.

AIDeep LearningStyleGAN
0 likes · 9 min read
AI-Powered High-Resolution Portrait Restoration Using StyleGAN and Face Parsing
Code DAO
Code DAO
Dec 12, 2021 · Artificial Intelligence

Lightning Flash 0.3 Introduces New Tasks, Visualization Tools, Data Pipelines, and Registry API

Lightning Flash 0.3 expands the PyTorch Lightning ecosystem with eight new computer‑vision and NLP tasks, modular API design, integrated model hubs, visualisation callbacks, customizable data‑source hooks, and a central registry for model backbones, all illustrated with concrete code examples.

Computer VisionDeep LearningLightning Flash
0 likes · 7 min read
Lightning Flash 0.3 Introduces New Tasks, Visualization Tools, Data Pipelines, and Registry API
Code DAO
Code DAO
Dec 11, 2021 · Artificial Intelligence

Nimble: A Lightweight Parallel GPU Scheduler Boosting Deep Learning Performance

The article analyzes how Nimble reduces GPU scheduling overhead and enables parallel execution through ahead‑of‑time scheduling and automatic multi‑stream assignment, achieving up to 22.3× inference speedup over PyTorch and significantly improving GPU utilization for deep learning workloads.

Deep LearningGPU schedulingParallel Execution
0 likes · 9 min read
Nimble: A Lightweight Parallel GPU Scheduler Boosting Deep Learning Performance
Kuaishou Large Model
Kuaishou Large Model
Dec 10, 2021 · Artificial Intelligence

How AI Restores Blurry Faces: Inside Kuaishou’s Y‑Tech High‑Definition Portrait Project

Image clarity impacts daily life, from personal memories to security, and Kuaishou’s Y‑Tech team tackles degradation by constructing paired low‑high quality datasets and a style‑based AI model that leverages facial masks to restore high‑definition portraits, preserving identity while enhancing detail.

AIComputer VisionDeep Learning
0 likes · 10 min read
How AI Restores Blurry Faces: Inside Kuaishou’s Y‑Tech High‑Definition Portrait Project
Code DAO
Code DAO
Dec 10, 2021 · Artificial Intelligence

Understanding Variational Autoencoders: From Dimensionality Reduction to Generative Modeling

This article explains the principles of variational autoencoders, starting with dimensionality reduction techniques such as PCA and standard autoencoders, highlighting their limitations for data generation, and then detailing VAE's regularized latent space, variational inference, re‑parameterization, and loss formulation.

Deep LearningGenerative ModelsKL divergence
0 likes · 18 min read
Understanding Variational Autoencoders: From Dimensionality Reduction to Generative Modeling
Code DAO
Code DAO
Dec 6, 2021 · Artificial Intelligence

Why So Many Optimizers? Core Algorithms Behind Neural Network Training

This article explains the fundamental gradient‑descent optimizers used in neural networks—SGD, Momentum, RMSProp, Adam and their variants—illustrates loss‑surface challenges such as local minima, saddle points and ravines, and shows how techniques like mini‑batching, momentum, adaptive learning rates and scheduling address these issues.

AdamDeep LearningMomentum
0 likes · 11 min read
Why So Many Optimizers? Core Algorithms Behind Neural Network Training
Code DAO
Code DAO
Dec 5, 2021 · Artificial Intelligence

Why DropBlock Outperforms Dropout as an Image Regularizer

This article demonstrates how to implement DropBlock in PyTorch, explains why Dropout fails on image data, details the gamma calculation and mask generation, and shows visual comparisons that illustrate the superiority of contiguous region dropping over random pixel dropout.

Computer VisionDeep LearningDropBlock
0 likes · 11 min read
Why DropBlock Outperforms Dropout as an Image Regularizer
Code DAO
Code DAO
Dec 5, 2021 · Artificial Intelligence

Why Neural Networks Need Batch Normalization: Principles and Mechanics

The article explains the principle behind Batch Normalization, why it is essential for training deep neural networks, how it standardizes activations, the role of learnable scale and shift parameters, the computation steps during training and inference, and discusses placement strategies within a model.

Batch NormalizationDeep LearningNeural Networks
0 likes · 9 min read
Why Neural Networks Need Batch Normalization: Principles and Mechanics