We deliver AI algorithm tutorials and the latest news, curated by a team of researchers from Peking University, Shanghai Jiao Tong University, Central South University, and leading AI companies such as Huawei, Kuaishou, and SenseTime. Join us in the AI alchemy—making life better!
This article dissects Optuna’s internal design by building three miniature versions (Minituna v1‑v3) that illustrate its main components, storage layer, sampling APIs, pruning mechanisms, and joint‑sampling concepts, while comparing them with Optuna’s full implementation.
The paper proposes a stand‑alone axial‑attention mechanism that converts 2‑D attention into 1‑D to lower computational cost while preserving global context, introduces position‑sensitive self‑attention, integrates it into Axial‑ResNet and Axial‑DeepLab, and demonstrates strong results on four large segmentation datasets.
This tutorial walks through constructing a movie recommendation engine by downloading a dataset, cleaning and de‑duplicating entries, encoding plot summaries into vectors with transformer models, and performing nearest‑neighbor searches using scikit‑learn, while handling misspellings with Levenshtein distance.
This article walks through building an image‑captioning system using a TensorFlow 2.0 encoder‑decoder with Bahdanau attention, covering dataset preparation, feature extraction with InceptionV3, model architecture, training with teacher forcing, and inference on the Flickr8K dataset.
This article explains the fundamentals of graph neural networks, covering graph concepts, node classification via neighborhood aggregation, message‑passing mechanics, mathematical notation, a full DGL‑PyTorch implementation on the Reddit dataset, and training results showing accuracy improvements up to 91 %.
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.
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.
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.
The article reviews the Context R-CNN paper, which introduces short‑term and long‑term memory banks and an attention mechanism to incorporate temporal context from multiple frames captured by a fixed camera, enabling robust detection of partially occluded, low‑light, distant, or background‑cluttered objects, and shows quantitative gains over standard Faster R‑CNN.
This article details a step‑by‑step implementation of visual servoing for a TurtleBot3 using a fisheye camera, ROS Melodic, OpenCV ArUco markers, camera calibration, A‑star path planning, and ROS publishers to move, avoid obstacles, and park precisely at a target pose.
