Ravid Shwartz-Ziv
Ravid Shwartz-Ziv
Work
Research
Writing & Podcast
Experience
Contact
Light
Dark
Automatic
Speech
S-JEPA: Soft Clustering Anchors for Self-Supervised Speech Representation Learning
Learning speech representations with soft predictive targets, preserving acoustic ambiguity while avoiding repeated offline reclustering.
Soft Clustering Anchors for Self-Supervised Speech Representation Learning in Joint Embedding Prediction Architectures
An earlier speech JEPA study using fixed soft clustering targets to stabilize representation learning.
JEPA as a Neural Tokenizer: Learning Robust Speech Representations with Density Adaptive Attention
Connecting JEPA representations to speech compression through density-adaptive attention and compact audio tokens.
Cite
×