DCT-based initialization and frequency truncation for self-attention improve accuracy and reduce overhead in Vision Transformers on standard benchmarks.
Fast fourier convolution.Advances in Neural Information Pro- cessing Systems, 33:4479–4488,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Discrete Cosine Transform Based Decorrelated Attention for Vision Transformers
DCT-based initialization and frequency truncation for self-attention improve accuracy and reduce overhead in Vision Transformers on standard benchmarks.