Paper Detail
Aditya Chetan, Eric Cai, Peeyush Kushwaha, Bharath Raj Nagoor Kani, Utkarsh Mall, Qianqian Wang, Noah Snavely, Bharath Hariharan
The emergence of Large Vision-Language Models (LVLMs) has significantly advanced video understanding capabilities. However, existing benchmarks focus predominantly on coarse-grained tasks such as action segmentation, classification, captioning, and retrieval. Furthermore, these benchmarks often rely on entities that can be easily identified verbally, like household objects, animals, human subjects, etc., limiting their applicability to complex, in-the-wild video scenarios. But, many applications such as furniture assembly, cooking, etc., require step-by-step fine-grained spatio-temporal understanding of the video, which is not sufficiently evaluated in current benchmarks. To address this gap, we introduce Flat-Pack Bench, a novel benchmark centered on furniture assembly tasks. Our benchmark evaluates LVLMs on nuanced tasks, including temporal ordering of assembly actions, temporal localization of assembly state, understanding part mating, and tracking, using multiple-choice questions paired with visual prompts highlighting relevant parts as references for fine-grained questions. Our experiments reveal that state-of-the-art LVLMs struggle significantly with fine-grained spatio-temporal reasoning, highlighting their limitations in effectively leveraging temporal information from videos, limited tracking ability, and understanding of spatial interactions like physical contact.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@misc{chetan2026flat,
title = {Flat-Pack Bench: Evaluating Spatio-Temporal Understanding in Large Vision-Language Models through Furniture Assembly},
author = {Aditya Chetan and Eric Cai and Peeyush Kushwaha and Bharath Raj Nagoor Kani and Utkarsh Mall and Qianqian Wang and Noah Snavely and Bharath Hariharan},
year = {2026},
abstract = {The emergence of Large Vision-Language Models (LVLMs) has significantly advanced video understanding capabilities. However, existing benchmarks focus predominantly on coarse-grained tasks such as action segmentation, classification, captioning, and retrieval. Furthermore, these benchmarks often rely on entities that can be easily identified verbally, like household objects, animals, human subjects, etc., limiting their applicability to complex, in-the-wild video scenarios. But, many applications},
url = {https://huggingface.co/papers/2605.21625},
keywords = {Large Vision-Language Models, video understanding, action segmentation, classification, captioning, retrieval, temporal ordering, temporal localization, part mating, tracking, spatio-temporal reasoning, code available, huggingface daily},
eprint = {2605.21625},
archiveprefix = {arXiv},
}
{}