Search Results for author:
Found 14 papers, 0 papers with code
Domino Saliency Metrics: Improving Existing Channel Saliency Metrics with Structural Information
The removal of these weight slices from a single layer causes mismatching number of feature maps between layers of the network.
Winograd Convolution for Deep Neural Networks: Efficient Point Selection
The choice of points impacts the numeric accuracy of the algorithm, but the optimal set of points for small convolutions remains unknown.
TASO: Time and Space Optimization for Memory-Constrained DNN Inference
We optimize the trade-off between execution time and memory consumption by: 1) attempting to minimize execution time across the whole network by selecting data layouts and primitive operations to implement each layer; and 2) allocating an appropriate workspace that reflects the upper bound of memory footprint per layer.
Composition of Saliency Metrics for Channel Pruning with a Myopic Oracle
In most cases our method finds better selections than even the best individual pruning saliency.
Performance-Oriented Neural Architecture Search
Hardware-Software Co-Design is a highly successful strategy for improving performance of domain-specific computing systems.
Taxonomy of Saliency Metrics for Channel Pruning
The result is that it is difficult to separate the effectiveness of the saliency metric from the wider pruning algorithm that surrounds it.
Explaining Reinforcement Learning to Mere Mortals: An Empirical Study
We present a user study to investigate the impact of explanations on non-experts' understanding of reinforcement learning (RL) agents.
Bonseyes AI Pipeline -- bringing AI to you. End-to-end integration of data, algorithms and deployment tools
In this work, we present a modular AI pipeline as an integrating framework to bring data, algorithms, and deployment tools together.
Scalar Arithmetic Multiple Data: Customizable Precision for Deep Neural Networks
Our approach yields very fast implementations of bit-precise custom DNN operations, which often match or exceed the performance of operations quantized to the sizes supported in native arithmetic.
Error Analysis and Improving the Accuracy of Winograd Convolution for Deep Neural Networks
We propose several methods for reducing FP error.
Optimal DNN Primitive Selection with Partitioned Boolean Quadratic Programming
We show experimentally that significant gains are possible versus the state of the art vendor libraries by using a principled analytic solution to the problem of layout selection in the presence of data format transformations.
Low-memory GEMM-based convolution algorithms for deep neural networks
We present two novel GEMM-based algorithms that require just $O(MHW)$ and $O(KW)$ additional space respectively, where $M$ is the number of channels in the result of the convolution.
Parallel Multi Channel Convolution using General Matrix Multiplication
A common approach to implementing convolutional layers is to expand the image into a column matrix (im2col) and perform Multiple Channel Multiple Kernel (MCMK) convolution using an existing parallel General Matrix Multiplication (GEMM) library.
