How reliable is an embedding of the seabed? Repeat-look measurement of frozen encoders on synthetic aperture sonar
cs.CV, cs.LG, stat.AP
Submitted: 2026-11-07
Updated: 2026-11-07
Terminology
Sources
- RankMe: Assessing the downstream performance of pretrained self-supervised representations by their rank
- BenthicDINO: Physics-Informed Self-Distillation for View-Invariant Side-Scan Sonar Representations
- Similarity of Neural Network Representations Revisited
- Mine-JEPA: In-Domain Self-Supervised Learning for Mine-Like Object Classification in Side-Scan Sonar
- Convergent Learning: Do different neural networks learn the same representations?
- Geometric Stability: The Missing Axis of Representations
- Self-Supervised Learning for Improved Synthetic Aperture Sonar Target Recognition
- DINOv3
- What Makes for Good Views for Contrastive Learning?
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