GitHub - Kavayk29/Speech-Recognition-using-TDNN-and-Data-Augmentation: Developed a speech recognition system using TDNN, preprocessing audio, extracting MFCC features, and training the model. Fine-tuning with augmented data (19,000 rows) improved accuracy from 9% to 80% training and 40% validation. Data augmentation proved crucial for enhancing model performance and generalization. Still working to increase the acc.

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
augmented_audio_dataset		augmented_audio_dataset
fine_tuned_model		fine_tuned_model
model		model
processed_recorded_audio		processed_recorded_audio
unprocessed_recorded_audio		unprocessed_recorded_audio
.gitignore		.gitignore
102103645_kavaykhurana.pdf		102103645_kavaykhurana.pdf
Converting_audio_into_smaller_audios.ipynb		Converting_audio_into_smaller_audios.ipynb
Preprocessing_data_and_Training_model.ipynb		Preprocessing_data_and_Training_model.ipynb
README.md		README.md
fine_tuning.ipynb		fine_tuning.ipynb
processed_data.csv		processed_data.csv
recorded_data_with_data_aug.csv		recorded_data_with_data_aug.csv
recorderd_data_with_data_aug_and_features.csv		recorderd_data_with_data_aug_and_features.csv

Provide feedback