saisjoshi Goto Github PK

📫 Reach me on [email protected]

I am Data Scientist and a Neural Engineering.

• Expert in applying deep learning to medical imaging, specializing in neuroimaging and radiology.
  
• Proficient in developing and deploying deep learning models (CNNs, ResNet, X-Net, Vision Transformers) for medical image analysis.
  
• Extensive experience in 3D medical image segmentation, template generation, and parcellation (PET, MRI).
  

• Strong programming foundation (Python, R, MATLAB, SQL) and machine learning expertise (PyTorch, Keras, scikit-learn).
  
• Proficiency in medical imaging software (FreeSurfer, PMOD, SPM, MonAI, SimpleITK).
  
• Skilled in data science techniques (statistical modeling, clustering, machine learning) for biomedical applications.
  

• Research and Development
  
• Proven ability to combine physics, bioengineering, and computer science for innovative solutions.
  
• Experience in building automated data preprocessing pipelines for 3D medical images.
  
• Strong leadership and collaboration skills, demonstrated in multi-university research projects.
  

• Brain-Computer Interface
• Deep Learning for Bioengineering
• Machine Learning for Neural Signal Processing
• Computer Vision and Image Segmentation for Biomedical Images
• Neural signals to Speech Conversion
• Adversarial Networks
• Neural Networks
• Natural Language Processing

Name	Task
Deep Learning I’m Something of a Painter Myself Kaggle Challenge	The goal of this competition is to implement GAN and evaluated on MiFID (Memorization-informed Frechet Inception ´ Distance), which is a modification from Frechet Inception Distance (FID). The smaller the MiFID is, the better the ´ generated images are.
Deep learning RSNA Screening Mammography Breast Cancer Detection Kaggle Challenge	The goal of this competition is to identify breast cancer. The task is to train your model with screening mammograms obtained from regular screening and work on improving the automation of detection in screening mammography may enable radiologists to be more accurate and efficient, improving the quality and safety of patient care. It could also help reduce costs and unnecessary medical procedures
Deep learning Automatic Speech recognition (ASR) Kaggle Challenge @ CMU	To perform automatic speech recognition using Levinshteiin Distance for evaluation. Obtained 5.04 accuracy
Deep learning Face Recognition and Verification Kaggle Challenge @ CMU	To implement CNN based architecture using ResNet, MobileNet, ConvNet for face recognition in images. Obtained 91.4 percent accuracy.
Deep learning Attention Based End to End Speech to Text Deep Neural Network Kaggle Challenge @ CMU	To implement sequence-to-sequence conversion, encoder-decoder architectures and attention.
Deep learning Utterance to Phoneme Mapping Kaggle Challenge @ CMU	To implement CTC (Connectionist Temporal Classification) and Decoding Strategies (Beam Search).
Deep learning Frame-Level Speech recognition Kaggle Challenge @ CMU	To create a multilayer perceptron for frame-level speech recognition. Obtained 87 percent accuracy
Machine learning Boston Dataset @ WPI	To analyze the data by performing different types of regression on the data using sklearn.
Machine learning MNIST Dataset @ WPI	To design and test KMean Clustering, PCA, decision tree, random forest, support vector machine (SVM), autoencoder classifier on the MNIST dataset To build a deep neural network with 5 hidden layers of 100 neurons each (with Xavier initialization, batch normalization, and ReLU activation function), and train it using Adam optimizer and early stopping on the MNIST dataset using sklearn.
Machine learning Diabetic Dataset @ WPI	To design and test lasso elastic net
Machine learning Olivetti Faces Database @ WPI	To design and test PCA and factor analysis on the Olivetti Faces Database.