About Me

Hello ! I'm Akash Patil,

I am a 2nd year master's student in Computer Engineering at Virginia Tech focusing in Software and Machine Intelligence. I am broadly interested in Machine Learning, Deep Learning, and Computer Vision. Currently, working on developing my skills in Deep Learning models and their applications. During Summer 2020, I worked as an intern at Qualcomm with Chipset Systems and Architecture team developing Machine Learning based optimization techniques for power grid analysis tool.

I graduated with a bachelor's degree in Electronics & Telecommunication from Fr. C. Rodrigues Institute of Technology, Mumbai in May 2018. I enjoy writing code in Python, C/C++, and Java.

My Resume.

Experience

Qualcomm

Machine Learning Intern
May 2020 – August 2020

Worked on chipset power grid generation and optimization using Machine Learning based techniques.
Developed Decision Tree and Random Forest based method for SMPS/LDO assignment to loads in the process of power grid generation.
Developed Greedy algorithm optimization method for reducing SMPS/LDO count in chipset power grids.
Developed a method to visualize and efficiently store raw (CSV) power grid data in JSON format.

Virginia Tech

Graduate Teaching Assistant, Advanced Machine Learning
January 2021 – May 2021
Graduate Teaching Assistant, Deep Learning
January 2020 – December 2020

Designed course assignments on CNN, RNN, and GANs in TensorFlow and PyTorch.
Helped and evaluated 100 students in course assignments and projects.

Eduvance

Industrial Training and Internship, Machine Learning
Jun 2018 - July 2018

Worked on various Scikit-learn Machine Learning techniques such as classification, regression, clustering, Dimensionality reduction, ensemble methods, and neural networks for data analysis.

Education

Virginia Tech

Master of Engineering, Computer Engineering
GPA: 3.9/4.0, August 2019 - May 2021

Relevant Coursework: Deep Learning, Adv Machine Learning, Computer Vision
Advanced Computer Vision, Applications of ML, Information Retrieval

Fr. C. Rodrigues Institute of Technology, Mumbai

Master of Engineering, Computer Engineering
GPA: 8.25/10.0, August 2014 - May 2018

Relevant Coursework: Microprocessors and Microcontrollers, Operating Systems, Object-oriented programming, Digital Image Processing

Projects

Few-Shot Image Classification using Meta-Learning

Virginia Tech | Deep Learning and Computer Vision | October 2020 – December 2020
  • Implemented the Model Agnostic Meta learning (MAML) algorithm in PyTorch for few-shot image classification.
  • Added GAN to the MAML model to generate data and help learn better decision boundaries for better performance.
  • Evaluated two models for few-shot image classification task on Omniglot and Mini ImageNet datasets.

Stand-Alone Self-Attention in Vision Models

Virginia Tech | Deep Learning and Computer Vision | March 2020 – May 2020 | Code: [github]
  • Reproduced results from above paper and performed experiments on SSD model with self-attention layers.
  • Implemented ResNet50 convolution network in PyTorch by replacing CNN layers with self-attention layers.
  • Analyzed effect of spatial extent, positional embedding, and attention type on ResNet50 attention model.
  • Implemented SSD object detection model by replacing set of convolution layers with self-attention layers.
  • Evaluated above model for performance, Accuracy (4.4% less), parameters (5.6% less), and Flop count (2% less).

SSD: Single Shot Multi-Box Detector

Virginia Tech, | Deep Learning and Computer Vision | October 2019 – December 2019 | Code: [github]
  • Implemented SSD object detection model using PyTorch framework.
  • Trained and tested SSD model on different datasets and measured accuracy to reproduce paper results.

Bio-medical Signal Processing using Machine Learning

Fr.C.R.I.T, | Machine Learning and Signal processing | August 2017 – March 2018
  • Developed a platform for ECG signal acquisition and processing using Python (Feature Extraction).
  • From extracted features ECG signals were classified using Multiclass SVM Machine Learning method.

Surveillance Robot

Fr.C.R.I.T, | Machine Learning | January 2017 – April 2017
  • Developed an object detection system for a robot using YOLO object detection algorithm in Python.

Skills

  • Programming Languages:
    Python (SciPy, NumPy, Pandas, NLTK, Dask), C, C++, Java
  • Machine Learning Frameworks:
    Scikit-learn, PyTorch, TensorFlow, Keras, OpenCV

Certifications

  • deeplearning.ai TensorFlow Developer Specialization, Coursera, Jan. 2020.
  • deeplearning.ai Deep Learning Specialization, Coursera, July 2018.