7 papers
Improved denoising diffusion probabilistic models with efficient non-diagonal covariance modeling
Rui Xia, Ayan Das, Artem Artemev +3
The sampling process of Denoising Diffusion Probabilistic Models (DDPMs) can be accelerated by leveraging second-order information in the form of approximations to the denoising po…
Text2Sign: A Single-GPU Diffusion Baseline for Text-to-Sign Language Video Generation
Ruize Xia
Sign language is a primary communication channel for millions of Deaf and hard-of-hearing people, yet text-to-signer video generation remains costly because video diffusion models…
Exploiting weight-space symmetries for approximating curvature
Artem Artemev, Rui Xia, Benjamin M. Boyd +4
Many machine learning techniques rely on approximating a loss function's curvature, but this is notoriously hard to do at the scale of modern deep networks. Surprisingly, no previo…
Matched-Learning-Rate Analysis of Attention Drift and Transfer Retention in Fine-Tuned CLIP
Ruize Xia
CLIP adaptation can improve in-domain accuracy while degrading out-of-domain transfer, but comparisons between Full Fine-Tuning (Full FT) and LoRA are often confounded by different…
Perturbation-efficient Zeroth-order Optimization for Hardware-friendly On-device Training
Qitao Tan, Sung-En Chang, Rui Xia +10
Zeroth-order (ZO) optimization is an emerging deep neural network (DNN) training paradigm that offers computational simplicity and memory savings. However, this seemingly promising…
Efficient Model Compression Techniques with FishLeg
Jamie McGowan, Wei Sheng Lai, Weibin Chen +7
In many domains, the most successful AI models tend to be the largest, indeed often too large to be handled by AI players with limited computational resources. To mitigate this, a…
Deep Transformers Thirst for Comprehensive-Frequency Data
Rui Xia, Chao Xue, Boyu Deng +2
Current researches indicate that inductive bias (IB) can improve Vision Transformer (ViT) performance. However, they introduce a pyramid structure concurrently to counteract the in…