Search Results for author: Vitaliy Chiley

Found 5 papers, 2 papers with code

RevBiFPN: The Fully Reversible Bidirectional Feature Pyramid Network

1 code implementation28 Jun 2022 Vitaliy Chiley, Vithursan Thangarasa, Abhay Gupta, Anshul Samar, Joel Hestness, Dennis Decoste

However, training them requires substantial accelerator memory for saving large, multi-resolution activations.

Ranked #309 on Image Classification on ImageNet (using extra training data)

General Classification Image Classification +2

Memory Efficient 3D U-Net with Reversible Mobile Inverted Bottlenecks for Brain Tumor Segmentation

no code implementations19 Apr 2021 Mihir Pendse, Vithursan Thangarasa, Vitaliy Chiley, Ryan Holmdahl, Joel Hestness, Dennis Decoste

The inverted residual bottleneck block uses lightweight depthwise separable convolutions to reduce computation by decomposing convolutions into a pointwise convolution and a depthwise convolution.

Brain Tumor Segmentation Tumor Segmentation

Adaptive Braking for Mitigating Gradient Delay

no code implementations2 Jul 2020 Abhinav Venigalla, Atli Kosson, Vitaliy Chiley, Urs Köster

Neural network training is commonly accelerated by using multiple synchronized workers to compute gradient updates in parallel.

Cannot find the paper you are looking for? You can Submit a new open access paper.