Return
Probabilistic image-based joint embedding predictive architecture
DOI:10.1016/j.patrec.2026.07.005.png)
Abstract
En 中文
• Proposed probabilistic approach optimizes the Joint Embedding Predictive Architecture.
• Alignment of target and predicted probability distributions.
• Improved performance in transfer learning and downstream tasks.
Keywords:
Computer vision
Self-Supervised learning
Representation learning
Journal
IF:
3.3
Papers:
7.9K
Citations:
1.6W
Organization
No organization information available
Cited Papers
No cited papers available

