BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Shengao, Wang, Wenqi, Wang, Zecheng, Whitton, Max, Wakeham, Michael, Chandra, Arjun, Huang, Joey, Zhu, Pengyue, Chen, Helen, Li, David, Li, Jeffrey, Li, Shawn, Zagula, Andrew, Zhao, Amy, Zhu, Andrew, Nakamura, Sayaka, Yamamoto, Yuki, Yokono, Jerry Jun, Mueller, Aaron, Plummer, Bryan A., Saenko, Kate, Saligrama, Venkatesh, Gong, Boqing
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914431109169152
author Wang, Shengao
Wang, Wenqi
Wang, Zecheng
Whitton, Max
Wakeham, Michael
Chandra, Arjun
Huang, Joey
Zhu, Pengyue
Chen, Helen
Li, David
Li, Jeffrey
Li, Shawn
Zagula, Andrew
Zhao, Amy
Zhu, Andrew
Nakamura, Sayaka
Yamamoto, Yuki
Yokono, Jerry Jun
Mueller, Aaron
Plummer, Bryan A.
Saenko, Kate
Saligrama, Venkatesh
Gong, Boqing
author_facet Wang, Shengao
Wang, Wenqi
Wang, Zecheng
Whitton, Max
Wakeham, Michael
Chandra, Arjun
Huang, Joey
Zhu, Pengyue
Chen, Helen
Li, David
Li, Jeffrey
Li, Shawn
Zagula, Andrew
Zhao, Amy
Zhu, Andrew
Nakamura, Sayaka
Yamamoto, Yuki
Yokono, Jerry Jun
Mueller, Aaron
Plummer, Bryan A.
Saenko, Kate
Saligrama, Venkatesh
Gong, Boqing
contents Early children's developmental trajectories set up a natural goal for sample-efficient pretraining of vision foundation models. We introduce BabyVLM-V2, a developmentally grounded framework for infant-inspired vision-language modeling that extensively improves upon BabyVLM-V1 through a longitudinal, multifaceted pretraining set, a versatile model, and, most importantly, DevCV Toolbox for cognitive evaluation. The pretraining set maximizes coverage while minimizing curation of a longitudinal, infant-centric audiovisual corpus, yielding video-utterance, image-utterance, and multi-turn conversational data that mirror infant experiences. DevCV Toolbox adapts all vision-related measures of the recently released NIH Baby Toolbox into a benchmark suite of ten multimodal tasks, covering spatial reasoning, memory, and vocabulary understanding aligned with early children's capabilities. Experimental results show that a compact model pretrained from scratch can achieve competitive performance on DevCV Toolbox, outperforming GPT-4o on some tasks. We hope the principled, unified BabyVLM-V2 framework will accelerate research in developmentally plausible pretraining of vision foundation models.
format Preprint
id arxiv_https___arxiv_org_abs_2512_10932
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models
Wang, Shengao
Wang, Wenqi
Wang, Zecheng
Whitton, Max
Wakeham, Michael
Chandra, Arjun
Huang, Joey
Zhu, Pengyue
Chen, Helen
Li, David
Li, Jeffrey
Li, Shawn
Zagula, Andrew
Zhao, Amy
Zhu, Andrew
Nakamura, Sayaka
Yamamoto, Yuki
Yokono, Jerry Jun
Mueller, Aaron
Plummer, Bryan A.
Saenko, Kate
Saligrama, Venkatesh
Gong, Boqing
Computer Vision and Pattern Recognition
Artificial Intelligence
Early children's developmental trajectories set up a natural goal for sample-efficient pretraining of vision foundation models. We introduce BabyVLM-V2, a developmentally grounded framework for infant-inspired vision-language modeling that extensively improves upon BabyVLM-V1 through a longitudinal, multifaceted pretraining set, a versatile model, and, most importantly, DevCV Toolbox for cognitive evaluation. The pretraining set maximizes coverage while minimizing curation of a longitudinal, infant-centric audiovisual corpus, yielding video-utterance, image-utterance, and multi-turn conversational data that mirror infant experiences. DevCV Toolbox adapts all vision-related measures of the recently released NIH Baby Toolbox into a benchmark suite of ten multimodal tasks, covering spatial reasoning, memory, and vocabulary understanding aligned with early children's capabilities. Experimental results show that a compact model pretrained from scratch can achieve competitive performance on DevCV Toolbox, outperforming GPT-4o on some tasks. We hope the principled, unified BabyVLM-V2 framework will accelerate research in developmentally plausible pretraining of vision foundation models.
title BabyVLM-V2: Toward Developmentally Grounded Pretraining and Benchmarking of Vision Foundation Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2512.10932