This article details the evolution of intelligent vending cabinets, comparing RFID, gravity, dynamic and static vision solutions, and explains how deep‑learning models, data pipelines, and system architectures enable high‑accuracy, low‑loss product detection and automated operations in modern unmanned retail.
This article introduces OpenCV.js, explains its basic concepts and demonstrates a complete workflow for detecting and correcting paper images in the browser using JavaScript, including matrix handling, resizing, filtering, edge detection, contour analysis, perspective transformation, and discusses challenges such as noise and incomplete edges.
This article reviews two recent computer‑vision approaches—DRRG and STKM—that combine CNN backbones with graph‑based relational reasoning and self‑attention to achieve state‑of‑the‑art detection of arbitrarily shaped text in images.
This article explains the challenges of scene text detection and introduces two state‑of‑the‑art models, ContourNet and CenterNet, detailing their architectural innovations, loss functions, and how they overcome issues like extreme aspect ratios and anchor‑based inefficiencies.
This article reviews two end‑to‑end deep‑learning approaches—DocUNet (CVPR 2018) and DewarpNet (ICCV 2019)—for correcting warped document images, detailing their network architectures, synthetic data generation, loss functions, experimental results, and the remaining challenges in document dewarping.
GoodFuture AI Institute secured four international titles at CVPR 2021—including Person In Context, UG²+, ETH‑XGaze, and ActivityNet—showcasing world‑class computer‑vision algorithms for human‑object interaction, low‑light face detection, gaze estimation, and active speaker detection, and highlighting their deployment in educational AI solutions.
The OSGAN method introduced by Alibaba’s Mama team and Prof. Song Ming‑Li merges generator and discriminator updates into a single stage, cutting GAN training time by roughly 1.5‑1.7× while maintaining performance, and is validated on symmetric and asymmetric DCGANs with open‑source code.
A Qingdao University student leveraged Intel Xeon SG1 GPU, OpenVINO and Mask R-CNN to transform existing parking‑lot cameras into an intelligent system that counts vehicles, detects pedestrians in blind spots, and issues real‑time safety alerts, showcasing a practical AI solution for child safety in crowded parking areas.
This article traces the origins, unique difficulties, method classifications, and current advancements of scene text detection, highlighting how AI has enabled computers to read images and the ongoing research to improve accuracy, speed, and multilingual support.
By deploying large‑scale visual AI—including a ResNet‑101 classifier, ArcFace‑trained matching features, clustering‑based sub‑category refinement, and product‑level image indexing—Xianyu’s marketplace dramatically improves image quality, removes duplicates, enhances search relevance and feed diversity, and filters non‑compliant content.
This article explains how a high‑traffic map service captures road features using client‑side computer‑vision models, details the deployment of many CNNs with the lightweight MNN engine on memory‑constrained devices, and shares practical memory‑saving techniques, inference scheduling, and error‑analysis methods.
At ICCV 2021 in Montreal, the LargeFineFoodAI workshop—co‑organized by Meituan Vision Intelligence Center, the Chinese Academy of Sciences, Beijing Zhiyuan and the University of Barcelona—will showcase state‑of‑the‑art fine‑grained food image research, feature invited speakers Jain, Aizawa and Radeva, and host a $12,000 prize challenge on Food2K across recognition and retrieval tracks.
This article demonstrates how to use Python's OpenCV and Pyzbar libraries to capture video from a webcam, decode barcodes in real time, display the results, and save captured frames, providing a practical guide for implementing barcode recognition in a retail checkout scenario.
The paper outlines an end‑to‑end pipeline that standardizes highly personalized food‑delivery dish names by combining rule‑based and BERT‑DSSM text synonym detection with EfficientNet image classification, constructing a multi‑level taxonomy that improves aggregation, supply‑demand analysis, recall ranking and merchant tagging.
This article details the algorithmic framework and baseline models for the 2021 Tencent Advertising Algorithm Competition, focusing on multimodal video ad parsing through temporal localization, scene segmentation, and multi-label classification to enhance advertising effectiveness and creative analysis.
This article reviews a GAN‑based deep learning method for removing motion blur from images, covering the problem definition, related work, the multi‑scale generator and discriminator architecture, loss functions, the GoPro dataset, and experimental results that demonstrate clear visual improvements.
The paper proposes BCNet, a lightweight bilayer instance segmentation network that explicitly models occluder and occludee relationships by treating each region of interest as two overlapping layers, achieving significant performance gains on COCO, COCOA and KINS datasets under heavy occlusion.
This article details Alibaba's AliMama brand risk identification system, covering the challenges of counterfeit detection, the construction of large‑scale brand datasets, the design of classification, logo detection, and variation models, their optimization, evaluation metrics, and future directions for AI‑driven brand protection.
This article reviews the CVPR2021 paper on Regressive Domain Adaptation (RegDA) for unsupervised keypoint detection, explaining its motivation, novel adversarial regression framework, sparse output-space modeling, min‑min training strategy, extensive experiments, and the resulting performance gains across multiple datasets.
The article presents Semantic Image Matting, a novel approach that incorporates 20 semantic Alpha pattern categories into the matting pipeline via semantic Trimap, region‑based classifiers, multi‑class discriminators, and learnable gradient loss, achieving state‑of‑the‑art results on multiple benchmarks.
The JD AI Institute and ICME2021 concluded their third face keypoint localization contest, emphasizing efficient masked‑face detection to aid COVID‑19 contact tracing, attracting top universities and tech firms, expanding data scale, and tightening model efficiency constraints to push the field forward.
The article explains Gaode Maps’ AR navigation design, detailing how environmental factors, spatial experience, color hierarchy, safety considerations, and competitor insights shape a six‑point design framework, and describes prototype testing, implementation strategies for overlapping alerts, and future prospects such as virtual road barriers and multimodal travel.
Camera-space hand mesh recovery (CMR) leverages semantic aggregation of 2D cues and adaptive 2D‑1D registration to predict absolute 3D hand pose and shape directly in camera coordinates, improving accuracy on benchmarks such as FreiHAND, RHD, and Human3.6M.
This article presents a method that replaces manual visual inspection with an automated YOLOv5‑based object detection pipeline to classify video frames as normal, colorful, or black screens, detailing data annotation, training, loss calculation, inference code, and showing a 97% accuracy improvement over ResNet.
Discover ten essential, free Python libraries for image processing—from scikit-image and NumPy to OpenCV-Python and Pycairo—each with resources, usage examples, and visual demonstrations, enabling you to manipulate, analyze, and transform images efficiently for computer vision and data science projects.
This article chronicles 58.com’s journey from a small brainstorming room to a full‑scale AI design platform, detailing the development of smart cutout, the BASNet segmentation model, custom loss functions, template editing, and the measurable business impact of the AI designer.
The talk reviews three major 2020 advances in computer vision—self‑supervised learning surpassing supervised pre‑training, the successful adoption of Transformer‑based attention models for detection and classification, and the emergence of Neural Radiance Fields for view synthesis—while highlighting related research from Microsoft Research Asia and the broader community.
The Youku AI team won the Media AI Alibaba Entertainment Video Object Segmentation Challenge by enhancing the STM model with a spatial‑constrained memory reader, ASPP‑HRNet refinement, ResNeSt‑101 backbone, and a multi‑stage training pipeline, while also devising an unsupervised framework that combines DetectoRS detection, HRNet mask refinement, STM‑based association, and key‑frame optimization to achieve 95.5% test score on a large, richly annotated video dataset.
Researchers from USTC and Kuaishou propose a frequency‑aware feature learning framework that combines a data‑driven adaptive frequency module with a novel single‑center loss, achieving state‑of‑the‑art performance on deepfake detection while addressing class‑distribution challenges.
This article details Kuaishou Y‑Tech's practical deployment of GAN‑based high‑precision face attribute editing—covering gender, age, hair, and expression transformations—for short‑video effects, discussing background, business applications, technical challenges, and solutions across data preparation, model training, and mobile deployment.
iQIYI’s OCR journey, explained by researcher Harlon, covers the evolution from separate detection and recognition pipelines to end‑to‑end models, key algorithms like CTPN, DB and CRNN, large‑scale simulated training, diverse video‑text applications, and future goals such as mobile deployment and tighter NLP integration.
This article describes how 58.com built an end‑to‑end deep‑learning watermark detection service, covering business needs, data collection and augmentation, model selection and iterative improvements (Faster‑RCNN, SSD, YOLOv3, anchor‑free methods), deployment results, and future research directions.
This tutorial walks through a project‑based OpenCV workflow that reads a digit template, preprocesses both template and credit‑card images, extracts individual numbers, matches them against the template, and finally overlays the recognized digits onto the original image, illustrating core computer‑vision techniques.
Amap’s visual‑technology pipeline automatically generates and updates POI names by crowdsourcing street‑level images, applying deep‑learning scene‑text recognition, dual‑branch classification of text attributes, and a BERT‑plus‑graph‑attention model that selects and orders recognized text, achieving about 95 % naming accuracy.
The article explores how modern AI technologies such as speech synthesis, natural language understanding, and the FastReID computer‑vision library enable realistic voice recreation and cross‑temporal dialogue, turning the emotional premise of the movie “Hi, Mom” into a tangible technical demonstration.
WeChat’s 8.0 update introduced an OCR pipeline that first quickly detects text in images, classifies the image type, applies a lightweight multi‑language detection network and a MobileNetV3‑based DBNet recognizer with a multi‑task CTC/Attention model, then merges results via a rule‑based layout analyzer to deliver accurate, well‑formatted extracted text across diverse languages and document types.
This article explains the popular Douyin AI effect “Ant Ya Hey”, showcases celebrity demos, provides a step‑by‑step guide using Avatarify and video editors, and delves into the underlying First‑Order Motion Model research that powers the realistic facial animation.
This article details Kuaishou Y‑tech’s Gorgeous beauty platform, covering traditional smoothing, advanced skin‑tone effects, AI‑driven blemish removal, clarity enhancement, local facial tuning, and the UNet‑based GorgeousGAN that delivers one‑click high‑definition beauty for live‑stream and short‑video applications.
This article explains a method for detecting video stutter by converting uploaded videos into frame sequences with ffmpeg, calculating pixel differences between consecutive frames, aggregating motion metrics, removing scene‑change effects, computing a dynamic factor, and outputting a binary result indicating the presence or absence of stutter.
This article presents Meitu Community's AI‑driven multimedia content analysis pipeline, covering short‑video classification, video fingerprinting, and OCR, detailing model choices, experimental results, and future directions for improving content audit, quality, tagging, and feature engineering.
JD Tech showcased a remarkable 21-paper presence at AAAI 2021, covering federated learning, spatio‑temporal AI, recommendation systems, computer vision, and causal learning, highlighting the company’s transition from research to real‑world AI applications across smart cities, retail, and risk management.
This guide compiles recommended books, courses, and open‑source projects across multimedia, artificial intelligence, computer vision, reinforcement learning, machine translation, knowledge graphs, Android, iOS, and frontend development to help newcomers and job seekers systematically deepen their technical expertise.
iQIYI’s AR+AI powered video interactive mini‑games blend a custom VideoAR engine with real‑time AI‑driven face and gesture detection, use lightweight Lua for game logic, and offer rapid hot‑updates, enabling diverse IP integrations that have attracted over a million participants and boosted viewer engagement.
The AMAP‑TECH competition challenged participants to infer real‑time road conditions from in‑vehicle video, prompting the authors to combine lane‑wise vehicle detection with LightGBM and later an end‑to‑end DenseNet‑GRU model, augment data, ensemble five networks, and achieve a 0.7237 F1 score while outlining future deployment and research directions.
This article details the Marlan Mountain Cup video point‑tracking challenge, describing the dataset, evaluation metrics, and a hybrid solution that combines SIFT‑based image registration with SiamMask tracking, along with extensive analysis and practical tricks for performance improvement.
This article explores the core portrait deformation techniques behind today’s beauty and body‑shaping effects—covering affine transforms, Moving Least Squares, triangulation, liquify, offset, 3D mesh, and deep‑learning approaches—detailing their principles, implementations, and visual results in live‑streaming and short‑video apps.
The article examines how iQIYI leverages AI across the video ecosystem—from intelligent material search, old‑film restoration, and voice cloning to virtual idols, XR production, and AI‑driven advertising—to boost creator efficiency, enhance user experience, and accelerate industry-wide digital transformation.
To tackle the AMAP‑TECH competition’s dynamic road‑condition classification from scarce, imbalanced vehicle‑video frames, the team combined YOLOv5 object detection, ResNeXt101‑based semantic embeddings, and engineered temporal detection statistics, feeding the fused features into a five‑fold LightGBM model that achieved top weighted‑F1 performance.
This article explains what OCR is, introduces the EasyOCR Python library, shows how to install it, walks through step‑by‑step usage with code examples, and summarizes the underlying deep‑learning techniques powering the library.
The article traces L4 autonomous-driving perception from early rule-based point-cloud methods through data-driven deep-learning models to emerging self-learning, multi-task systems, and highlights four key hurdles—model generalization and explainability, robust multi-sensor fusion, real-time compute limits, and proper uncertainty handling—calling for integrated AI, engineering, and data solutions.
An open‑source PyTorch‑based project enables users to remove unwanted objects from videos simply by drawing a bounding box around them, offering a practical demo, step‑by‑step instructions, and a GitHub repository with over 2 k stars.
On January 9, 2021, Meituan hosted an unmanned‑delivery technical salon in Beijing where experts presented cutting‑edge AI research—including the CenterMask instance‑segmentation method, 3D geometry‑aware camera localization, multi‑agent trajectory prediction with attention‑based spatio‑temporal graphs, real‑time stereo visual‑inertial odometry calibration, and self‑supervised depth‑pose learning for dynamic scenes.
This article outlines SuNing's unmanned store technology, comparing its data-driven, product selection, and customer experience advantages over traditional shops, and detailing AI-powered applications such as face detection, target tracking, image recognition, and 3D reconstruction that enable 24‑hour service, intelligent merchandising, and precise customer analytics.
This article details the evolution of WeChat’s Identify product, explaining its end‑to‑end image recognition pipeline—including object detection, multi‑label classification, mobile‑side detection, large‑scale retrieval, unsupervised clustering, and system architecture—while showcasing various application scenarios such as product, plant, and landmark recognition.
FrankMocap, an open‑source AI algorithm from Facebook AI Research and HKU, delivers simultaneous 3D full‑body and hand pose estimation from a single monocular video, runs at about 9.5 FPS on a RTX 2080, and includes easy installation steps, code examples, and links to its GitHub repository and paper.
This article introduces a Java OCR project for ID cards that integrates OpenCV, JavaCPP, and Tess4J to perform image preprocessing, region cropping, and character recognition without requiring OpenCV installation, and details its features, encountered issues, system requirements, updates, and source repository.
iQIYI’s AI‑driven UI element recognizer uses a YOLOv3 model trained on thousands of mobile and PC screenshots to locate obfuscated controls across diverse devices, integrating predictions into its IDE for reliable automation of React Native, Flutter, H5 and mini‑program interfaces.
This article provides a comprehensive overview of Generative Adversarial Networks, covering their original formulation, training dynamics, loss functions, major variants such as DCGAN and WGAN, and practical implementations for image‑to‑image translation, style transfer, and handwriting font synthesis at Laiye Technology.
Y‑Tech’s in‑house keypoint detection system powers Kuaishou’s beauty and effect filters across live streaming, video creation, and editing by leveraging lightweight deep‑learning models, extensive multi‑scenario data collection, and specialized handling of occlusion, enabling real‑time, robust facial and body landmark tracking on diverse mobile devices.
This article reviews common image similarity detection techniques—including hash-based methods (aHash, pHash, dHash), histogram comparison, feature matching with ORB and SIFT/SURF, bag‑of‑words with K‑Means, and CNN‑based VGG16 features—detailing their algorithms, Python implementations, performance characteristics, and practical considerations.
Li Yongxiang, technical director at Suning Tech, outlines how AI-driven visual unmanned stores and integrated big‑data, cloud, and edge computing solutions enable low‑cost digital transformation across thousands of retail outlets, improving shopper experience, inventory management, and operational efficiency.
This article presents Meituan's AI-driven short video analysis workflow, covering industry trends, multi‑label video classification, intelligent cover selection, and video generation techniques, while discussing challenges, model building, label expansion, continuous data iteration, and future outlook for video AI in local services.
The presentation explains AR fundamentals, distinguishes it from AI and VR, and details a shoe‑try‑on system that captures 30 fps video, uses AI key‑point detection and pose estimation to overlay 3D shoe models—created via manual, scanning, or photogrammetry methods—rendered with GPU pipelines and PBR, enhanced by green‑screen occlusion and shadow techniques, earning positive audience feedback.
A Scottish football match turned comedic when an AI camera misidentified a linesman's shiny bald head as the ball, highlighting challenges in automated sports broadcasting and prompting fan backlash over the technology's shortcomings.
This article describes how Didi’s vehicle‑vision team built an ultra‑low‑power ADAS solution on the JueShi dash‑cam platform, using lightweight detection models, temporal fusion, camera‑calibration techniques and data‑driven optimization to cut rear‑end collision rates by over 11% and improve overall traffic safety.
DiDi’s ultra‑low‑power JueShi ADAS combines lightweight vision models, temporal‑fusion Kalman filtering, and camera‑calibration techniques to deliver real‑time forward‑collision warnings and brake‑light alerts, cutting rear‑end crashes by over 11% and overall accidents by 9% through continuous edge‑AI learning.
This article introduces the fundamentals of object detection, explains the latest YOLOv4 and YOLOv5 architectures, and details the essential components—including data preparation, regularization, backbone, neck, and prediction innovations—along with label smoothing and advanced loss functions for improved detection performance.
This article explains the fundamental principles of image recognition, covering how images are converted to numeric arrays, processed by scanning matrix blocks, and matched against patterns to identify objects such as text, faces, cats, dogs, or mice.
iQIYI’s ACM Multimedia‑accepted paper unveils iCartoonFace, the world’s largest manually annotated cartoon‑face dataset—over 5,000 characters and 400,000 real‑scene images—accompanied by a semi‑automatic collection pipeline and multi‑person training framework, now powering AI services, large‑scale contests and accelerating cartoon‑character recognition research.
Didi’s COVID‑19 mask‑detection system, built on a DFS‑based face detector and an attention‑enhanced ResNet‑50 mask classifier achieving over 99.5 % accuracy, has been deployed in vehicles, open‑sourced, and complemented by top‑ranked results in international visual AI contests, including first place in driver‑gaze prediction and podium finishes in emotion recognition and model‑compression challenges.
This article reviews Kuashou Y‑Tech’s ECCV‑2020 paper on monocular depth estimation, detailing its novel GCB‑SAB network, new HC‑Depth dataset, specialized loss functions and edge‑aware training, and demonstrates superior performance on NYUv2, TUM and real‑world mobile AR applications.
To enhance object detection robustness, the article discusses various data augmentation methods—including rotation, flipping, random cropping, scaling, color jitter, blurring, transparency adjustment, and image partitioning—providing code examples and illustrating their impact on model performance with before‑and‑after results.
SuNing’s fourth‑generation fully digital visual unmanned store combines 3D reconstruction, AI‑driven perception, and modular hardware‑software design to achieve real‑time, all‑scene, all‑time, all‑digital analysis of people, goods, and spaces, enabling precise offline marketing and scalable retail digitization.
Suning’s lecture series details the three‑decade evolution of retail, the company’s 30‑year digital transformation, and the technical architecture of its fourth‑generation fully digital visual unmanned stores that leverage AI, computer vision, and big‑data analytics to revolutionize in‑store operations and customer experience.
This article presents ZTFace, an end‑to‑end face recognition solution that integrates face detection, alignment, feature embedding, verification, anti‑spoofing and attribute recognition using deep learning, details its backbone networks, loss functions, training datasets, experimental results on WIDER FACE and LFW, and demonstrates acceleration with TensorRT.
This guide explains image matching techniques, covering template matching with OpenCV, various matching methods, SIFT feature extraction and description, FLANN-based nearest neighbor matching, homography estimation, practical challenges, and a brief overview of YOLO training, providing code examples and visual illustrations.
The Boundary Content Graph Neural Network (BC‑GNN) introduces a bipartite‑graph framework that jointly refines start/end boundary probabilities and segment‑content confidence, enabling more precise temporal action proposals and achieving state‑of‑the‑art results on ActivityNet‑1.3 and THUMOS14.
Amap uses advanced scene text recognition combining detection and recognition modules, deep learning, data synthesis, and result fusion to automate map data production, achieving state-of-the-art performance and automating the majority of POI and road updates, significantly reducing labor costs.
This article explains how Amap leverages scene text recognition to automate map data production, detailing the evolution from traditional image algorithms to deep‑learning models, the current detection and recognition framework, performance results, and future research directions for handling blur, data scarcity, and semantic understanding.
This article explains how Gaode Maps leverages traditional and deep‑learning based scene text recognition techniques—including character detection, sequence models, data synthesis, and multi‑stage frameworks—to automate POI and road data production with high precision and speed.
The talk explains how Alibaba’s Entertainment Content Operation Platform applies a Partial‑ReID algorithm to overcome the challenges of person re‑identification in heavily edited video content, enabling accurate cross‑shot character matching, richer appearance data, and metrics such as presence, interaction, and storyline for improved video quality assessment.
This survey reviews video action recognition, comparing 3D convolutional networks that jointly model spatial‑temporal cues but are computationally heavy with 2D‑based approaches like TSM and TIN that embed temporal shifts efficiently, and emphasizes how large‑scale pre‑training markedly improves performance despite limited labeled data.
This article explains the principles and implementation of face detection based on statistical methods, detailing Haar feature types, integral image computation, feature normalization, cascade classifiers, and provides step‑by‑step OpenCV code examples for static images, eye detection, and real‑time webcam detection.
The AMAP-TECH algorithm competition invites participants to develop AI models that analyze in-vehicle video sequences to determine dynamic road conditions, offering detailed dataset specifications, evaluation metrics, expert judges, schedule, and prize information for researchers in computer vision and traffic analytics.
This technical talk introduces video object segmentation tasks, reviews leading datasets and state-of-the-art deep learning models, and shares practical network design rules and performance‑boosting techniques, presented by Prof. Wang Xinggang as part of Alibaba's MEDIA AI challenge series.
Alibaba Amap’s AMAP‑TECH competition invites participants to develop AI computer‑vision models that classify real‑time road conditions—smooth, slow, or congested—from short sequences of dash‑cam images, using a labeled dataset of 1,500 training sequences and a weighted F1‑score evaluation, with cash prizes up to ¥60,000.
This comprehensive review explains visual object tracking in computer vision, covering its definition, core sub‑problems of candidate generation, feature extraction, and decision making, system architecture, motion, feature and observation models, algorithm classifications, evaluation metrics, datasets, and recent research trends.
In the fourth Alibaba Digital Media Technology Night Talk, algorithm engineer Liu Xiaolong presents an overview of video‑based temporal event detection, covering its problem background, representative prior works, and the latest research advances within the MEDIA AI Algorithm Challenge series.
This article explains the fundamentals of virtual reality and its distinction from augmented reality, describes VR hardware, outlines depth‑estimation and eye‑tracking methods such as projection, Hough transform, AdaBoost and sample matching, discusses Sobel edge detection, and explores the importance of audio, haptic feedback, and immersive VR applications in education.
This article introduces image matching fundamentals, covering template matching methods, feature-based approaches such as SIFT and FLANN, their implementation details, matching rules, homography transformation, and practical considerations, providing a comprehensive overview for computer vision applications.
The JD ARVR technology department announced two peer‑reviewed papers—one on a novel defocus blur detection network published in Transaction on Multimedia and another on a transductive relation‑propagation network for few‑shot learning accepted at IJCAI 2020—highlighting their advanced AI research and future AR‑VR ecosystem plans.
Gaode’s automated production line combines deep‑learning image recognition, GPS‑enhanced location services, image differencing with semantic filtering, and standardized data‑fusion to continuously refresh China’s national base map, cutting manual effort and costs while delivering real‑time, high‑quality map updates for road traffic infrastructure.
This article explains the AR Cut & Paste prototype by Cyril Diagne, detailing its three‑module architecture, the underlying BASNet and U²‑Net vision models, and provides a step‑by‑step guide—including code snippets and GitHub links—to set up the mobile app, local server, and Photoshop integration.
This tutorial demonstrates how to use Python and the PaddlePaddle deep‑learning platform to automatically remove backgrounds from multiple photos in one step, covering installation, verification, and a concise five‑line code example for batch human segmentation.
ChineseOCR Lite is an open‑source, ultra‑lightweight OCR solution that supports vertical Chinese text, runs on Linux/macOS via ncnn inference, and packs detection, recognition, and angle classification models into a total of just 17 MB, offering fast and accurate scene‑text processing.
Didi’s DGF‑M model, a mask‑aware face‑recognition AI, combines multi‑task training and synthetic data to detect masks with under 0.1 % miss rate and verify identities with up to 99.5 % pass rate at a 0.1 % false‑acceptance rate, and is deployed for driver verification, offered through the Didi Cloud API marketplace, and released as an open‑source solution to aid pandemic‑era security.
In a live Q&A, Alibaba Amap’s chief scientist Ren Xiaofeng explained how computer‑vision algorithms underpin high‑precision map creation, AR navigation, visual localization and sensor fusion, discussed current hardware limits, deep‑learning bottlenecks, 5G’s role, edge‑cloud cooperation, and offered career advice for transitioning researchers.
Jason Mayes, a Google web engineer, open‑sourced a TensorFlow.js demo that removes people from live webcam video in real time using a lightweight MobileNet model, with only about 200 lines of code, and provides GitHub and CodePen links for experimentation.
The iCartoonFace Challenge invites participants to develop efficient algorithms for detecting and recognizing cartoon faces using large, meticulously annotated datasets—50,000 images for detection and nearly 390,000 for recognition—while meeting strict model size and latency limits and submitting detailed methods and code.
This article surveys state‑of‑the‑art deep‑learning techniques for scene text detection and recognition in e‑commerce poster images, detailing models such as CTPN, TextBoxes, SegLink, EAST, and end‑to‑end frameworks, and discusses their architectures, strengths, limitations, and future challenges.
