What Really Works for Universal Multimodal Embeddings? A Controlled Study of Contrastive Training Strategies
cs.CV
Submitted: 2026-10-10
Updated: 2026-10-10
Terminology
Sources
- Discovering Global False Negatives On the Fly for Self-supervised Contrastive Learning
- LoRA: Low-Rank Adaptation of Large Language Models
- MMEB-V3: Measuring the Performance Gaps of Omni-Modality Embedding Models
- E5-V: Universal Embeddings with Multimodal Large Language Models
- VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks
- Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking
- VLM2Vec-V2: Advancing Multimodal Embedding for Videos, Images, and Visual Documents
- Contrastive Learning with Hard Negative Samples
- EVA-CLIP: Improved Training Techniques for CLIP at Scale
- Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
- Improve Multi-Modal Embedding Learning via Explicit Hard Negative Gradient Amplifying
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models