Can I read Multi-Label Image Classification with C-Tran on EtoBox?

Multi-Label Image Classification with C-Tran by zhangzhengyi443 is a document available to read on EtoBox.

What is Multi-Label Image Classification with C-Tran about?

The document proposes a new framework called Classification Transformer (C-Tran) for multi-label image classification that leverages transformers to exploit dependencies among visual features and labels. C-Tran consists of a transformer encoder trained to predict target labels given input labels with some masked, and visual features from a CNN. During training, C-Tran learns to reconstruct a partial set of labels given randomly masked input label embeddings. During inference, C-Tran can predict labels using

Author

zhangzhengyi443

Language

EN