AVGen-Bench: A Task-Driven Benchmark for Multi-Granular Evaluation of Text-to-Audio-Video Generation
This addresses the need for better evaluation tools in media creation for researchers and developers, though it is incremental as it builds on existing multimodal assessment methods.
The paper tackles the problem of fragmented evaluation in text-to-audio-video generation by introducing AVGen-Bench, a task-driven benchmark with 11 real-world categories, and reveals a significant gap between strong aesthetics and weak semantic reliability, including failures in text rendering and musical pitch control.
Text-to-Audio-Video (T2AV) generation is rapidly becoming a core interface for media creation, yet its evaluation remains fragmented. Existing benchmarks largely assess audio and video in isolation or rely on coarse embedding similarity, failing to capture the fine-grained joint correctness required by realistic prompts. We introduce AVGen-Bench, a task-driven benchmark for T2AV generation featuring high-quality prompts across 11 real-world categories. To support comprehensive assessment, we propose a multi-granular evaluation framework that combines lightweight specialist models with Multimodal Large Language Models (MLLMs), enabling evaluation from perceptual quality to fine-grained semantic controllability. Our evaluation reveals a pronounced gap between strong audio-visual aesthetics and weak semantic reliability, including persistent failures in text rendering, speech coherence, physical reasoning, and a universal breakdown in musical pitch control. Code and benchmark resources are available at http://aka.ms/avgenbench.