Task Me Anything

TL;DR

Task-Me-Anything is a flexible benchmark generation engine that creates tailored, large-scale multimodal evaluation datasets to assess the specific strengths and weaknesses of large multimodal language models across various tasks.

Contribution

It introduces a novel, extendable system for generating customized multimodal benchmarks, addressing the challenge of selecting appropriate evaluations for specific applications.

Findings

Open-source MLMs excel in object and attribute recognition.

Models show weaknesses in spatial and temporal understanding.

Larger models generally perform better, with some exceptions.

Abstract

Benchmarks for large multimodal language models (MLMs) now serve to simultaneously assess the general capabilities of models instead of evaluating for a specific capability. As a result, when a developer wants to identify which models to use for their application, they are overwhelmed by the number of benchmarks and remain uncertain about which benchmark's results are most reflective of their specific use case. This paper introduces Task-Me-Anything, a benchmark generation engine which produces a benchmark tailored to a user's needs. Task-Me-Anything maintains an extendable taxonomy of visual assets and can programmatically generate a vast number of task instances. Additionally, it algorithmically addresses user queries regarding MLM performance efficiently within a computational budget. It contains 113K images, 10K videos, 2K 3D object assets, over 365 object categories, 655 attributes, and 335 relationships. It can generate 750M image/video question-answering pairs, which focus on evaluating MLM perceptual capabilities. Task-Me-Anything reveals critical insights: open-source MLMs excel in object and attribute recognition but lack spatial and temporal understanding; each model exhibits unique strengths and weaknesses; larger models generally perform better, though exceptions exist; and GPT4o demonstrates challenges in recognizing rotating/moving objects and distinguishing colors.