Ask, Pose, Unite: Scaling Data Acquisition for Close Interactions with Vision Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bravo-Sánchez, Laura, Heo, Jaewoo, Weng, Zhenzhen, Wang, Kuan-Chieh, Yeung-Levy, Serena
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929522378539008
author Bravo-Sánchez, Laura
Heo, Jaewoo
Weng, Zhenzhen
Wang, Kuan-Chieh
Yeung-Levy, Serena
author_facet Bravo-Sánchez, Laura
Heo, Jaewoo
Weng, Zhenzhen
Wang, Kuan-Chieh
Yeung-Levy, Serena
contents Social dynamics in close human interactions pose significant challenges for Human Mesh Estimation (HME), particularly due to the complexity of physical contacts and the scarcity of training data. Addressing these challenges, we introduce a novel data generation method that utilizes Large Vision Language Models (LVLMs) to annotate contact maps which guide test-time optimization to produce paired image and pseudo-ground truth meshes. This methodology not only alleviates the annotation burden but also enables the assembly of a comprehensive dataset specifically tailored for close interactions in HME. Our Ask Pose Unite (APU) dataset, comprising over 6.2k human mesh pairs in contact covering diverse interaction types, is curated from images depicting naturalistic person-to-person scenes. We empirically show that using our dataset to train a diffusion-based contact prior, used as guidance during optimization, improves mesh estimation on unseen interactions. Our work addresses longstanding challenges of data scarcity for close interactions in HME enhancing the field's capabilities of handling complex interaction scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2410_00309
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Ask, Pose, Unite: Scaling Data Acquisition for Close Interactions with Vision Language Models
Bravo-Sánchez, Laura
Heo, Jaewoo
Weng, Zhenzhen
Wang, Kuan-Chieh
Yeung-Levy, Serena
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Social dynamics in close human interactions pose significant challenges for Human Mesh Estimation (HME), particularly due to the complexity of physical contacts and the scarcity of training data. Addressing these challenges, we introduce a novel data generation method that utilizes Large Vision Language Models (LVLMs) to annotate contact maps which guide test-time optimization to produce paired image and pseudo-ground truth meshes. This methodology not only alleviates the annotation burden but also enables the assembly of a comprehensive dataset specifically tailored for close interactions in HME. Our Ask Pose Unite (APU) dataset, comprising over 6.2k human mesh pairs in contact covering diverse interaction types, is curated from images depicting naturalistic person-to-person scenes. We empirically show that using our dataset to train a diffusion-based contact prior, used as guidance during optimization, improves mesh estimation on unseen interactions. Our work addresses longstanding challenges of data scarcity for close interactions in HME enhancing the field's capabilities of handling complex interaction scenarios.
title Ask, Pose, Unite: Scaling Data Acquisition for Close Interactions with Vision Language Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2410.00309