Update README and adding images

desinurch · Jul 27, 2020 · 310cbd2 · 310cbd2
1 parent ff48dba
commit 310cbd2
Show file tree

Hide file tree

Showing 145 changed files with 21 additions and 20,459 deletions.
diff --git a/A Comparative Inference Evaluation of DNN Compression Methods-Nurchalifah.pdf b/A Comparative Inference Evaluation of DNN Compression Methods-Nurchalifah.pdf
diff --git a/README.md b/README.md
@@ -1,7 +1,25 @@
-# Research and Development Project HBRS
+# Research and Development Project
 ## A Comparative Study of Sparsity Methods in Deep Neural Network for Faster Inference
 
-Code and documentation for research and development project with a topic in Deep Neural Network Compression as partial fulfillment in Masters of Autonomous Systems program.
+	Code and documentation for research and development project with a topic in Deep Neural Network Compression as partial fulfillment in Masters of Autonomous Systems program.
+
+## Overview
+
+	Comparison of compression methods in Deep Learning for image classification task. Comparison is done in terms of speed using the backbone of [MLMark benchmark](https://www.eembc.org/mlmark/). Compression methods observed are as follows:
+
+	![Compression Methods](/imgs/methods-compression.png)
+
+## Description
+### Dataset
+	Dataset used for comparison is [CIFAR-10](https://www.cs.toronto.edu/~kriz/cifar.html) to mimic real-life situations.
+
+### Model Architecture
+	Dataset are processed using the network of ResNet-56 and ResNet-110 with pre-activations. In model distillation mode, both of the network act as a teacher which knowledge are transferred to student networks; ResNet-1, ResNet-10, and ResNet-20
+
+## Results
+![Speedup vs Compression](/imgs/speedup_vs_compression.png)
+ ---------------------------------------------------------------------------
+![Accuracy vs Speedup](/imgs/accuracyLoss_vs_sppedUp.png)
 
 ### Repository structure:
 - L1 norm pruning: cloned from https://github.com/Eric-mingjie/rethinking-network-pruning/tree/master/cifar/l1-norm-pruning with minor modifications. Based on the implementation of the paper [Pruning Filters For Efficient ConvNets](https://arxiv.org/pdf/1608.08710.pdf)
@@ -11,4 +29,4 @@ Code and documentation for research and development project with a topic in Deep
 	- FitNets implementation. Cloned from https://github.com/AberHu/Knowledge-Distillation-Zoo with modifications. Based on the implementation of the paper [FitNets: Hints for Thin Deep Nets](https://arxiv.org/pdf/1412.6550.pdf)
 - Low Rank Approximations [Caffe: Convolutional Architecture for Fast Feature Embedding](https://arxiv.org/abs/1408.5093)
 - Quantizations (https://github.com/eladhoffer/convNet.pytorch/blob/master/models/modules/quantize.py)
-- Report of research and development project
+- Results presentations
diff --git a/imgs/accuracyLoss_vs_sppedUp.png b/imgs/accuracyLoss_vs_sppedUp.png
diff --git a/imgs/methods-compression.png b/imgs/methods-compression.png
diff --git a/imgs/speedup_vs_compression.png b/imgs/speedup_vs_compression.png
diff --git a/src/FitNets/train_at.py b/src/FitNets/train_at.py