Skip to search formSkip to main contentSkip to account menu
EIE: Efficient Inference Engine on Compressed Deep Neural Network
TLDR
An energy efficient inference engine (EIE) that performs inference on this compressed network model and accelerates the resulting sparse matrix-vector multiplication with weight sharing and is 189x and 13x faster when compared to CPU and GPU implementations of the same DNN without compression.
View on ACM
Exploring the Regularity of Sparse Structure in Convolutional Neural Networks
TLDR
This paper quantitatively measure the trade-off between sparsity regularity and prediction accuracy, providing insights in how to maintain accuracy while having more a more structured sparsity pattern.
View PDF on arXiv
Efficient Sparse-Winograd Convolutional Neural Networks
TLDR
Two modifications to Winograd-based CNNs are proposed to enable these methods to exploit sparsity, including moving the ReLU operation into the Winog Rad domain to increase the sparsity of the transformed activations and prune the weights in the Winogsrad domain to exploit static weight sparsity.
View PDF on arXiv
Exploring the Granularity of Sparsity in Convolutional Neural Networks
TLDR
This analysis, based on the framework of a recent sparse convolutional neural network (SCNN) accelerator, demonstrates that coarse-grained sparsity saves 30% – 35% of memory references compared with fine-graining sparsity.
View on IEEE
Deep compression and EIE: Efficient inference engine on compressed deep neural network
This article consists only of a collection of slides from the author's conference presentation.
View on IEEE
Head pose Estimation Using Convolutional Neural Networks
TLDR
This project explores using convolutional neural networks that take RGB image as input to estimate the head pose using regression as the estimation approach, and proves that proper regularization strength could prevent overfitting thus reduce test loss.
Accurate Prediction = > Large Model = > More Memory Reference = > High Power
GoGoGo : Improving Deep Neural Network Based Go Playing AI with Residual Networks
TLDR
AlphaGo, a Go-playing AI built by Google DeepMind, used a new approach of combining deep neural networks with tree search to solve the Go playing problem and defeated the 18-time Go world champion Lee Sedol.
By clicking accept or continuing to use the site, you agree to the terms outlined in our Privacy Policy, Terms of Service, and Dataset License