This tutorial explains how to explore the COCO keypoint detection annotation files, describes their JSON structure and fields, and provides step‑by‑step Python code using json, Pillow, and matplotlib to load images, extract keypoints, and draw both points and skeletal connections for visual analysis.
This article provides a detailed, English-language walkthrough of the HRNet source code, covering how the COCO keypoint dataset is built, the online data‑augmentation techniques applied during training, and the end‑to‑end training and inference procedures for human pose estimation.
This article presents a comprehensive overview of ModelScope's computer‑vision models, detailing visual detection and keypoint solutions—including VitDet, YOLOX, res2net, HRNet, and 3D pose models—their architectures, performance highlights, real‑world applications, and future development plans.
The article details how AI‑based image animation and face‑swapping decompose video motion into zero‑order rigid and first‑order affine components via Taylor expansion, using unsupervised U‑Net keypoint extraction, sparse-to-dense optical flow conversion, and dense motion networks that learn masks for region‑wise rigidity and non‑rigid deformation.
This article reviews the end‑to‑end pipeline for real‑time document corner detection on mobile devices, breaks down the keypoint detection workflow into image processing, encoding, network modeling and decoding, compares heatmap‑based and fully‑connected approaches, introduces a differentiable DSNT decoding method with unbiased coordinate transformations, and presents experimental results and conclusions on its effectiveness and limitations.
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.
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 paper presents a deep learning-based approach for selecting cover images in Autohome forum articles, employing Faster R-CNN for object detection, Mask R-CNN for human keypoint detection, and MobileNetV2 for attribute recognition, achieving an overall accuracy of 81.5%.
The Firefly team secured first place in the AI Challenger human skeletal keypoint detection contest by employing a top‑down approach that combines Faster R‑CNN for person detection, a multi‑scale region‑feature fusion strategy, varied Gaussian and binary supervision, OKS‑NMS post‑processing, and extensive experiments demonstrating the impact of input size, supervision radius, and model stacking.
