contrastive loss
Contrastive loss is a loss function used to distinguish between similar and dissimilar pairs of samples, often employed in tasks involving representation learning and unsupervised learning. This method helps to embed similar instances close together in the feature space while pushing apart distinct instances.
- CSPCL: Category Semantic Prior Contrastive Learning for Deformable DETR-Based Prohibited Item Detectors
- Compress & Cache: Vision token compression for efficient generation and retrieval
- Contrastive Consolidation of Top-Down Modulations Achieves Sparsely Supervised Continual Learning
- Contrastive Learning with Data Misalignment: Feature Purity, Training Dynamics and Theoretical Generalization Guarantees
- How Different from the Past? Spatio-Temporal Time Series Forecasting with Self-Supervised Deviation Learning
- Orthogonal Contrastive Learning for Multi-Representation fMRI Analysis
- Prot2Text-V2: Protein Function Prediction with Multimodal Contrastive Alignment
- Scaling Laws for Robust Comparison of Open Foundation Language-Vision Models and Datasets
- Towards Reliable Identification of Diffusion-based Image Manipulations
- VITRIX-UniViTAR: Unified Vision Transformer with Native Resolution