AI Engineering/Human Understanding 2024.06.08

[Motion Generation] MotionLLM: Understanding Human Behaviors from Human Motions and Videos

https://lhchen.top/MotionLLM/ MotionLLM: Understanding Human Behaviors from Human Motions and VideosVideos comprehension of models. The results show good performance of MotionLLM on captioning, spatial-temporal comprehension, and reasoning. The comparison with Video-Chat and Video-LLaVA shows good sequentiality and direction comprehension of MotionLLM. (lhchen.top  MotionLLM은 MotionLCM, HumanTOM..

AI Engineering/Human Understanding 2023.11.15

[Human Pose Estimation] A Dataset of Relighted 3D Interacting Hands

Paper : https://arxiv.org/pdf/2310.17768.pdf Project Page: https://mks0601.github.io/ReInterHand/ Re:InterHand Dataset A Dataset of Relighted 3D Interacting Hands (NeurIPS 2023 Datasets and Benchmarks Track) mks0601.github.io Meta Reality Labs Research 에서 나온 데이터세트 Re:InterHand 데이터이다. 두 손이 상호작용 하는 문제는 self-similarity, complicated articulations, and occlusions of hands 문제 때문에 가장 분석하기 어려운 task 중 하나..

AI Engineering/Human Understanding 2023.08.18

[Paper Review] Effective Whole-body Pose Estimation with Two-stages Distillation

Paper : https://arxiv.org/abs/2307.15880 Effective Whole-body Pose Estimation with Two-stages Distillation Whole-body pose estimation localizes the human body, hand, face, and foot keypoints in an image. This task is challenging due to multi-scale body parts, fine-grained localization for low-resolution regions, and data scarcity. Meanwhile, applying a highly e arxiv.org GitHub : https://github...

AI Engineering/Human Understanding 2023.05.26

[Paper Review] MobileHumanPose : Toward real-time 3D human pose estimation in mobile devices

이번에 읽게된 논문은 CVPR 2021에 소개된 MobileHumanPose: Toward real-time 3D human pose estimation in mobile devices 입니다. 참고로 카이스트에서 게재한 논문이며, 3D Human Pose Estimation 모델을 Mobile Device에서 작동 될 수 있도록 경량화 한 논문입니다. Contribution 본 논문에서의 주요 Contribution은 아래와 같습니다. 기존 3D HPE 방법들이 높은 computing cost + 정확도에 초점을 맞췄던 것에 비해 본 논문에서는 모바일 기반 모델 효율성을 다룸 MobileNet v2 수정, parametric activation function, Skip concatenation (U-..

AI Engineering/Human Understanding 2023.05.23

[Human Pose Estimation] AdaFuse: Adaptive Multiview Fusion for Accurate Human Pose Estimation in the Wild

GitHub : https://github.com/zhezh/occlusion_person GitHub - zhezh/occlusion_person: A dataset for multiview 3D human pose estimation with detailed occlusion labels, powered by Unr A dataset for multiview 3D human pose estimation with detailed occlusion labels, powered by UnrealCV - GitHub - zhezh/occlusion_person: A dataset for multiview 3D human pose estimation with detaile... github.com Paper ..

AI Engineering/Human Understanding 2023.03.31

[Paper Review] Adaptive Wing Loss for Robust Face Alignment via Heatmap Regression

GitHub : https://github.com/protossw512/AdaptiveWingLoss Paper : https://arxiv.org/abs/1904.07399 이번에 리뷰할 논문은 ICCV 2019에 소개되었던 Adaptive Wing Loss 논문입니다. Introduction 본 논문에서는 face alignment (= facial landmark localization)문제에 대해 heatmap regression 방법을 사용할 때 효과적인 Adaptive Wing loss를 소개합니다. 이 loss function은 foreground pixel과 background pixel의 불균형 문제를 해결하기 위해 등장하였고, 여기서 Adaptive가 의미하는 바는 background ..

AI Engineering/Human Understanding 2023.03.16

[Paper Review] RTMPose: Real-Time Multi-Person Pose Estimation based on MMPose

Paper : https://arxiv.org/pdf/2303.07399v1.pdf GitHub : https://github.com/open-mmlab/mmpose/tree/1.x/projects/rtmpose GitHub - open-mmlab/mmpose: OpenMMLab Pose Estimation Toolbox and Benchmark. OpenMMLab Pose Estimation Toolbox and Benchmark. Contribute to open-mmlab/mmpose development by creating an account on GitHub. github.com 오늘 리뷰할 논문은 Shanghai AI Lab에서 소개한 "RTMPose: Real-Time Multi-Perso..

AI Engineering/Human Understanding 2023.02.23

[Paper Review] Keypoint-wise Adaptive Loss for Whole-Body Human Pose Estimation

Paper : https://www.researchgate.net/publication/368336170_Keypoint-wise_Adaptive_Loss_for_Whole-Body_Human_Pose_Estimation Introduction 오늘 읽어볼 논문은 AAAI 2023에 소개될 Keypoint-wise Adaptive Loss for Whole-Body Human Pose Estimation 라는 논문입니다. NHN Cloud 분들이 논문을 냈네요. 아직 코드는 공개가 안된 듯 합니다. 이 논문은 dense 및 coarse keypoints의 mixed characteristic을 분석하여 whole-body human pose estimation을 수행합니다. whole body를 추정하는..

AI Engineering/Human Understanding 2023.02.20

[Paper Review] MDM: Human Motion Diffusion Model

Paper : https://arxiv.org/pdf/2209.14916.pdfGitHub : https://github.com/GuyTevet/motion-diffusion-model GitHub - GuyTevet/motion-diffusion-model: The official PyTorch implementation of the paper "Human Motion Diffusion Model"The official PyTorch implementation of the paper "Human Motion Diffusion Model" - GitHub - GuyTevet/motion-diffusion-model: The official PyTorch implementation of the paper ..

AI Engineering/Human Understanding 2023.02.20

[Paper Review] SimCC: a Simple Coordinate Classification Perspective for Human Pose Estimation

Paper : https://arxiv.org/abs/2107.03332 GitHub : https://github.com/leeyegy/SimCC GitHub - leeyegy/SimCC: [ECCV'2022 Oral] PyTorch implementation for: SimCC: a Simple Coordinate Classification Perspective for H [ECCV'2022 Oral] PyTorch implementation for: SimCC: a Simple Coordinate Classification Perspective for Human Pose Estimation (http://arxiv.org/abs/2107.03332). Old name: SimDR - GitHub -..