Perception Test: A Diagnostic Benchmark for Multimodal Video Models

Viorica Patraucean*, Dima Damen, Andrew Zisserman, Joao Carriera

*Corresponding author for this work

Research output: Contribution to conferenceConference Paper

2 Downloads (Pure)

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, BEiT-3, 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 significant gap in performance (91.4% vs 45.8%), suggesting that there is significant room for improvement in multimodal video understanding. Dataset, baselines code, and challenge server are available at https://github.com/deepmind/perception_test
Original languageEnglish
Pages1-27
Number of pages27
Publication statusPublished - 16 Dec 2023
EventConference on Neural Information Processing Systems - Ernest N. Morial Convention Center, New Orleans, United Kingdom
Duration: 10 Dec 202316 Dec 2023

Conference

ConferenceConference on Neural Information Processing Systems
Country/TerritoryUnited Kingdom
CityNew Orleans
Period10/12/2316/12/23

Fingerprint

Dive into the research topics of 'Perception Test: A Diagnostic Benchmark for Multimodal Video Models'. Together they form a unique fingerprint.

Cite this