SEPS: Semantic-enhanced Patch Slimming Framework for fine-grained cross-modal alignment
cs.CV, cs.AI, cs.MM
Submitted: 2025-11-03
Updated: 2026-11-16
Code: https://github.com/Sweet4tars/seps
Terminology
Sources
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- VSE++: Improving Visual-Semantic Embeddings with Hard Negatives
- Aligning Information Capacity Between Vision and Language via Dense-to-Sparse Feature Distillation for Image-Text Matching
- The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
- DecAlign: Hierarchical Cross-Modal Alignment for Decoupled Multimodal Representation Learning
- HALoGEN: Fantastic LLM Hallucinations and Where to Find Them
- ReCLIP: A Strong Zero-Shot Baseline for Referring Expression Comprehension
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs