CVAICLOct 17, 2025

OmniVinci: Enhancing Architecture and Data for Omni-Modal Understanding LLM

arXiv:2510.15870v233 citationsh-index: 45Has Code
Originality Incremental advance
AI Analysis

This work addresses the need for efficient multi-modal AI in applications like robotics and medical AI, though it appears incremental with architectural and data improvements.

The paper tackled the problem of building an omni-modal LLM for cross-modal understanding by introducing OmniVinci, which achieved performance gains such as +19.05 on DailyOmni and used 6 times fewer training tokens than a baseline.

Advancing machine intelligence requires developing the ability to perceive across multiple modalities, much as humans sense the world. We introduce OmniVinci, an initiative to build a strong, open-source, omni-modal LLM. We carefully study the design choices across model architecture and data curation. For model architecture, we present three key innovations: (i) OmniAlignNet for strengthening alignment between vision and audio embeddings in a shared omni-modal latent space; (ii) Temporal Embedding Grouping for capturing relative temporal alignment between vision and audio signals; and (iii) Constrained Rotary Time Embedding for encoding absolute temporal information in omni-modal embeddings. We introduce a curation and synthesis pipeline that generates 24M single-modal and omni-modal conversations. We find that modalities reinforce one another in both perception and reasoning. Our model, OmniVinci, outperforms Qwen2.5-Omni with +19.05 on DailyOmni (cross-modal understanding), +1.7 on MMAR (audio), and +3.9 on Video-MME (vision), while using just 0.2T training tokens - a 6 times reduction compared to Qwen2.5-Omni's 1.2T. We finally demonstrate omni-modal advantages in downstream applications spanning robotics, medical AI, and smart factory.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes