4 papers
TraceSampler: Making Sampling Errors Predictable and Correctable for Accelerated Generative Sampling
George Matthew, Guoqiang Zhang, Kenta Niwa
Accurate sampling from continuous-time generative models can require many model evaluations. Low-step sampling reduces cost, but coarse updates introduce endpoint errors that can a…
Confusion-Aware Spectral Regularizer for Long-Tailed Recognition
Ziquan Zhu, Gaojie Jin, Hanruo Zhu +11
Long-tailed image classification remains a long-standing challenge, as real-world data typically follow highly imbalanced distributions where a few head classes dominate and many t…
High Quality Diffusion Distillation on a Single GPU with Relative and Absolute Position Matching
Guoqiang Zhang, Kenta Niwa, J. P. Lewis +2
We introduce relative and absolute position matching (RAPM), a diffusion distillation method resulting in high quality generation that can be trained efficiently on a single GPU. R…
On Exact Bit-level Reversible Transformers Without Changing Architectures
Guoqiang Zhang, J. P. Lewis, W. B. Kleijn
Various reversible deep neural networks (DNN) models have been proposed to reduce memory consumption in the training process. However, almost all existing reversible DNNs either re…