Speech Lab of LTRC, IIITH is a very active lab. It is a place of constant innovation and engaged research. I am currently pursuing my honours under Dr Chiranjeevi Yarra. I am presently working in spoken language forensics and informatics (SFLI). More specifically, I am working on the Acoustic Unit Discovery problem. For the honours project, I have to be good with all the essential machine learning tools, maybe sometimes even tweak them to create better labels.

I am expecting to do good in the honours since this would be a stepping stone for me to get into a good MS program. I wish to master various machine learning topics through research, hard work and persistency.

I have learnt various machine learning concepts during the semester's honours project. I learnt ...

1. Using TensorFlow and Building various custom neural networks

2. Cluster Analysis - (K-Means, GMM)

3. Regression

4. Experimentation and Performance Evaluation

5. Post and Pre processing (Practical Difficulties)

6. Image Processing

7. Ensemble Methods, Decision Trees

I have built code for K-Means, GMM, ANNs, CNNs, on MNIST data. I have built neural networks for a custom 2-class classification of the syllable stress (for a given encoding of a speech signal of a syllable, the neural network has to classify whether it has been stressed or not.), done post-processing on it, achieved an accuracy of 88%.