🔎 Open for Opportunities!

• 🌟 Seeking Summer 2025 Internships in AI/ML Research
• 💡 Interest Areas: Computer Vision, Self-Supervised Learning, Foundation Models
• 🤝 Open to Research Collaborations in Deep Learning & Planetary Science
• 📫 Reach out: tpanambur@umass.edu

I'm an AI Research Scientist and Ph.D. candidate at UMass Amherst, specializing in machine learning, computer vision, and self-supervised learning. Currently, I'm advancing the field of self-supervised learning and multimodal AI to support NASA's Mars mission analysis through the Remote Hyperspectral Observers (RHO) group. I've worked with SETI and Nokia Bell Labs on cutting-edge AI models as a research intern for planetary exploration and industry applications.

• Self-supervised learning & deep clustering
• Representation learning & deep texture analysis
• Multimodal foundation models
• Generative AI
• Computer vision

• CVPRw 2022: Self-Supervised Learning for Martian Terrain Categorization
• IGARSS 2021-2023: Multiple publications on deep clustering and texture recognition
• Pioneering work in applying deep learning to planetary science datasets

• Frontier Development Labs (SETI): Developed transformer-based multimodal models for planetary data analysis
• Nokia Bell Labs: Advanced active learning strategies for efficient model training
• NASA Collaboration: Improved Mars terrain analysis through novel deep learning approaches

• MIT COVID-19 Challenge Winner (2020)

  • Track: Hospital-assets coordination, distribution, and management
  • Developed innovative solutions for the automation of healthcare resource optimization
  • View Project Details

• Robert Bosch India - 2nd Prize (2016)

  • Technical Paper Presentation at INSCRIBE 2016
  • Project: "Agrosquad - Agricultural Automation"

languages = ['Python', 'C++', 'C', 'MATLAB']

tools = {
    'Deep Learning': ['PyTorch', 'TensorFlow', 'Keras'],
    'Data Science': ['Pandas', 'NumPy', 'Scikit-learn'],
    'Computer Vision': ['OpenCV'],
    'Cloud': ['Microsoft Azure'],
    'NLP': ['NLTK', 'Spacy'],
    'Big Data': ['Dask']
}

• Developed novel self-supervised clustering algorithms
• Implemented deep texture encoding with triplet networks
• Achieved 15% improvement in clustering accuracy
• Collaborated with planetary geology experts

• 3D Reconstruction: Shape and pose prediction from 2D images
• Domain Adaptation: Cross-domain classification for Mars rover datasets
• Video Colorization: Deep learning-based B&W video colorization
• Speech Recognition: End-to-end ASR pipeline with deep neural networks

• Ph.D. in Electrical and Computer Engineering, UMass Amherst (Expected Dec 2025)
• M.S. in Electrical and Computer Engineering, UMass Amherst (2020)

• 💼 LinkedIn
• 🌐 Portfolio
• 📧 Email

• Advancing self-supervised learning techniques for Martian Terrain Classification
• Developing multimodal foundation models for finding correlations between hyperspectral datasets
• Improving SOTA generalized category discovery