https://github.com/JanMarcelKezmann/TensorFlow-Advanced-Segmentation-Models
Deep High-Resolution Representation Learning for Visual Recognition
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
Pose Estimation is a computer vision task where the goal is to detect the position and orientation of a person or an object. Usually, this is done by predicting the location of specific keypoints like hands, head, elbows, etc. in case of Human Pose Estimation.
A common benchmark for this task is MPII Human Pose
<span style="color:grey; opacity: 0.6">( Image credit: Real-time 2D Multi-Person Pose Estimation on CPU: Lightweight OpenPose )</span>
관련 논문과 평가 결과에 연결된 구현입니다. 출처에서 확인된 저자 구현에는 공식 코드 표시가 있습니다.
Deep High-Resolution Representation Learning for Visual Recognition
High-Resolution Representations for Labeling Pixels and Regions
SuperGlue: Learning Feature Matching with Graph Neural Networks
CrowdPose: Efficient Crowded Scenes Pose Estimation and A New Benchmark
MuCAN: Multi-Correspondence Aggregation Network for Video Super-Resolution
Deep Alignment Network: A convolutional neural network for robust face alignment
Coherent Reconstruction of Multiple Humans from a Single Image
Unsupervised Deep Homography: A Fast and Robust Homography Estimation Model
OpenPose: Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields
Learning Temporal Coherence via Self-Supervision for GAN-based Video Generation
Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields
Keep it SMPL: Automatic Estimation of 3D Human Pose and Shape from a Single Image
M3D-RPN: Monocular 3D Region Proposal Network for Object Detection
ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation
MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion
The Pose Knows: Video Forecasting by Generating Pose Futures
Hand Keypoint Detection in Single Images using Multiview Bootstrapping
3D Human Motion Estimation via Motion Compression and Refinement
Distribution-Aware Coordinate Representation for Human Pose Estimation
Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.