DONDAPATI SUDEEP

Recent graduate with strong research, technical, and problem-solving skills. Detail-oriented with a proven ability to quickly grasp new concepts. Committed to contributing to organizational success through effective knowledge application and skill enhancement.

• English
• Telugu
• Hindi

• Music
• Badminton

• Worked on a project to recognize human emotions from speech using deep learning techniques.
• Collected and preprocessed audio data from standard datasets like RAVDESS and SAVEE.
• Extracted key speech features such as MFCC, pitch, and energy for model input.
• Designed and trained deep learning models including CNNs, Transformers, and hybrid architectures.
• Focused on classifying emotions.
• Improved model accuracy through tuning and experimentation with various architectures.
• Evaluated performance using metrics like accuracy, confusion matrix, and loss curves.

• Built a Convolutional Neural Network (CNN) model to classify handwritten digits using the MNIST dataset.
• Preprocessed image data for training and testing the model.
• Designed the model using convolutional, pooling, and fully connected layers to improve learning and accuracy.
• Trained the model to recognize digits from 0 to 9.
• Evaluated the model's performance using accuracy and loss metrics.
• Gained practical understanding of how deep learning can be applied to image classification tasks.