GitHub - atul-khobragade/Deep-Learning-DL-model-optimization-for-Lightweight-Keyword-Spotting.-: Two DL model Implementation with high accuracy for Google Speech Command Dataset namely DNN and CNN.

atul-khobragade / Deep-Learning-DL-model-optimization-for-Lightweight-Keyword-Spotting.- Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Two DL model Implementation with high accuracy for Google Speech Command Dataset namely DNN and CNN.

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
B20CS027_ESC7.zip		B20CS027_ESC7.zip
B20CS027_ESC7_Report.pdf		B20CS027_ESC7_Report.pdf
ELU_Activation_Function_With_Optimizer_Experiment.ipynb		ELU_Activation_Function_With_Optimizer_Experiment.ipynb
Embedded_project_MODEL1_CNN.ipynb		Embedded_project_MODEL1_CNN.ipynb
Embedded_project_MODEL2_DNN.ipynb		Embedded_project_MODEL2_DNN.ipynb
README.txt		README.txt
ReLu_Activation_Function_With_Optimizer_Experiment.ipynb		ReLu_Activation_Function_With_Optimizer_Experiment.ipynb

Repository files navigation

ESC7 - Project code
Contributers - Karan Singh Kushwah & Atul Khobragade

Four Code files are provided - 
1. Embedded Project Model1 CNN
2. Embedded Project Model2 DNN
3.ELU Activation Function With Optimizer Experminet
4.ReLu Activation Function With Optimizer Experminet

How to run the code ?
1. Open Google colab on your browser
2. Sign in with your google account
3. Load the file you want to run
4. Select runtime(use default runtime)
5. Select run all cells


Features - 
1.Two DL model Implementation with high accuracy for Google Speech Command Dataset namely DNN and CNN.
2. Pre-processing of dataset in MFCC format.
3. Experiment with different layers and Parameters available(it really took long time )
4. Final model with best parameters and got the best accuracy and minimum loss with CNN and DNN.
5. Report is provided with full analysis of Project



Thank you :)