Institution profile

ASMPT

Industry researchasia · hk
Official website
Research library1linked papers
Opportunities0open roles
Selected work

Representative Papers

Decorrelation Speeds Up Vision Transformers

Oct 16, 2025

Vision Transformers (ViTs) pretrained via Masked Autoencoders (MAEs) achieve strong performance under low-label regimes, yet their high computational cost hinders industrial deployment. To address this, we propose **Selective Decorrelation Backpropagation (DecorrBP)**—a lightweight optimization technique that imposes layer-wise gradient covariance constraints exclusively within the MAE encoder, enhancing gradient propagation efficiency and convergence speed while preserving training stability. Evaluated on ImageNet-1K, DecorrBP reduces pretraining time by 21.1% and carbon emissions by 21.4%. On downstream ADE20K semantic segmentation, it improves mIoU by 1.1 points; consistent gains are also observed on industrial datasets. Crucially, DecorrBP is the first method to integrate gradient decorrelation into the MAE training framework without modifying model architecture or loss functions—enabling efficient, low-carbon, and high-performance ViT pretraining.

0 citationsRead paper
Recent publications

Latest Papers

Decorrelation Speeds Up Vision Transformers

Oct 16, 2025

Vision Transformers (ViTs) pretrained via Masked Autoencoders (MAEs) achieve strong performance under low-label regimes, yet their high computational cost hinders industrial deployment. To address this, we propose **Selective Decorrelation Backpropagation (DecorrBP)**—a lightweight optimization technique that imposes layer-wise gradient covariance constraints exclusively within the MAE encoder, enhancing gradient propagation efficiency and convergence speed while preserving training stability. Evaluated on ImageNet-1K, DecorrBP reduces pretraining time by 21.1% and carbon emissions by 21.4%. On downstream ADE20K semantic segmentation, it improves mIoU by 1.1 points; consistent gains are also observed on industrial datasets. Crucially, DecorrBP is the first method to integrate gradient decorrelation into the MAE training framework without modifying model architecture or loss functions—enabling efficient, low-carbon, and high-performance ViT pretraining.

0 citationsRead paper