Perception Test: A Diagnostic Benchmark for Multimodal Video Models
Abstract
We propose a novel multimodal video benchmark - the Perception Test - to evaluate the perception and reasoning skills of pre-trained multimodal models (e.g. Flamingo, SeViLA, or GPT-4). Compared to existing benchmarks that focus on computational tasks (e.g. classification, detection or tracking), the Perception Test focuses on skills (Memory, Abstraction, Physics, Semantics) and types of reasoning (descriptive, explanatory, predictive, counterfactual) across video, audio, and text modalities, to provide a comprehensive and efficient evaluation tool. The benchmark probes pre-trained models for their transfer capabilities, in a zero-shot / few-shot or limited finetuning regime. For these purposes, the Perception Test introduces 11.6k real-world videos, 23s average length, designed to show perceptually interesting situations, filmed by around 100 participants worldwide. The videos are densely annotated with six types of labels (multiple-choice and grounded video question-answers, object and point tracks, temporal action and sound segments), enabling both language and non-language evaluations. The fine-tuning and validation splits of the benchmark are publicly available (CC-BY license), in addition to a challenge server with a held-out test split. Human baseline results compared to state-of-the-art video QA models show a substantial gap in performance (91.4% vs 46.2%), suggesting that there is significant room for improvement in multimodal video understanding. Dataset, baseline code, and challenge server are available at https://github.com/deepmind/perception_test
Cite
@article{arxiv.2305.13786,
title = {Perception Test: A Diagnostic Benchmark for Multimodal Video Models},
author = {Viorica Pătrăucean and Lucas Smaira and Ankush Gupta and Adrià Recasens Continente and Larisa Markeeva and Dylan Banarse and Skanda Koppula and Joseph Heyward and Mateusz Malinowski and Yi Yang and Carl Doersch and Tatiana Matejovicova and Yury Sulsky and Antoine Miech and Alex Frechette and Hanna Klimczak and Raphael Koster and Junlin Zhang and Stephanie Winkler and Yusuf Aytar and Simon Osindero and Dima Damen and Andrew Zisserman and João Carreira},
journal= {arXiv preprint arXiv:2305.13786},
year = {2023}
}
Comments
37th Conference on Neural Information Processing Systems (NeurIPS 2023) Track on Datasets and Benchmarks