OmniCompass: PolyU launches a holistic benchmark for omni-modal video grounding and QA
VCLab-HKPU · hf · 2026-10-16
Hong Kong Polytechnic University's VCLab, with OPPO Research Institute and others, released OmniCompass, a holistic benchmark for omni-modal video grounding and understanding.
- Motivation: existing video benchmarks cover only slices of the evaluation space, limiting systematic diagnosis of MLLMs
- Two tracks: OmniCompass-VL for tasks requiring visual evidence, OmniCompass-AV for tasks requiring joint audio-visual evidence
- Each track includes three complementary tasks: temporal grounding (TG), spatial grounding (SG), and question answering (QA)
- Available on GitHub and Hugging Face
A more complete yardstick for systematically evaluating video-cap multimodal models.
More from Multimodal
- vLLM officially supports Muse Spark Open: multimodal, 1M context, agentic — vllm_project · 2026-10-16
- InfoVis: 9B unified model tops open-source slide generation and editing, trained on 13M records — Jiahao Wang · 2026-10-16
- RepForcing: feature-space supervision improves few-step autoregressive video distillation — zju · 2026-10-16
- SCOPE: semantic span-aware GRPO objective improves T2I prompt rewriting — university-of-rochester · 2026-10-16
- Orbia benchmark asks: do generated videos form a consistent 3D world? — tencent · 2026-10-16
- 3DCodeVerse: USC's ecosystem generates verifiable 3D worlds as code, with a 2M-sample dataset — UniversityofSouthernCalifornia · 2026-10-16