Ziyang Wu

representation learning (4) token compression (2) sparse rate reduction (2) transformer architecture (2) continual learning (1) catastrophic forgetting (1) dimensionality reduction (1) incremental learning (1) vision transformer (1) few-shot learning (1) model selection (1) network architecture (1) variational inference (1) transfer learning (1) deformable convolution (1) metric learning (1) embedding learning (1) question answering (1) feature extraction (1) zero-shot learning (1)