sooryamsharma / adv-pose-estimation

Adversarial Learning for Human Pose Estimation: Generative Adversarial Deep Convolutional Networks to attain Structure-awareness in Human Pose Estimation | MS Thesis

Hi there 👋 I'm Suryam 👨‍💻

🔧 Senior Backend Engineer at Ernst & Young | 🖼️ Computer Vision Enthusiast

My work in the tech industry has honed my skills in backend development, but here on GitHub, my passion for Computer Vision takes center stage.

🌟 Areas of Passion:

• 💡 Facial Sentiment Analysis: Understanding the complexities of human emotions through technology.
• 🏃 Activity Recognition: Using AI to interpret and anticipate human actions.

📌 On GitHub, my code meets vision 🔭. Dive into my projects and let's exchange ideas.

• 🌱 I’m currently working on my research on Facial Sentiment Analysis

---

📃 My Resume

Education

• 📖 Master of Science - Computer Science (GPA:3.63/4.00)
  📆 August 2018 - May 2021
  🏤 University of Texas at Arlington - Arlington TX, USA

• 📖 Bachelor of Engineering - Computer Engineering
  📆 June 2013 - August 2017
  🏤 University of Pune - Pune MH, India

Work Experience

• 💼 Sr. Backend Engineer
  📆 March 2022 - Present
  📍 Ernst & Young - Dallas TX, USA | Banking Transformation Project
  - Collaborated on architecting a domain-driven party service application integrating legacy MDM with modern cloud-based banking platforms.
  - Revolutionized customer onboarding for instant account setup and access to cutting-edge financial tools.
  - Leveraged Spring Boot for our party service application, optimized for cloud performance via AWS containerization.
  - Developed scalable RESTful APIs connecting seamlessly with top cloud-based financial platforms.
  - Employed Amazon MSK for reliable cross-domain messaging, leveraging Apache Kafka.
  - Incorporated Splunk for enhanced system logging, monitoring, and real-time troubleshooting insights.
  - Designed intuitive dashboards for instant insights into app performance and user behavior.
  - Made bi-monthly client visits to ensure clear communication and alignment with client vision.
  - Excelled in translating complex user requirements into clear designs and actionable plans.
  - Contributed to thorough project documentation including Swagger, Confluence, architecture diagrams, data models, and more.
  - Assisted in designing robust automated Newman test scenarios for consistent software quality.
  - Volunteered to automate a Python script for another team, extracting crucial metadata from ~90K queries.
  - Aided in engagement economics tasks, creating Tableau dashboards and handling monthly invoices.

• 💼 Software Developer
  📆 September 2021 - March 2022
  📍 Capital One Financial - Plano TX, USA
  - Engaged in backend development, specifically focusing on Kafka streams and REST APIs, to enhance auto loan loss mitigation efforts.
  - Worked hand-in-hand with the Data Science team, developing Java integrations for ML models geared towards repossession strategies.
  - Spanned the entire SDLC: from requirements and design brainstorming to agile coding.
  - Oversaw thorough end-to-end QA and performance tests, ensuring top-tier production quality.
  - Amplified unit tests using JUnit, boasting a stellar code coverage surpassing 95%.
  - Devised user acceptance test scenarios for data streams & APIs via Cucumber Gherkins.

• 💼 Graduate Research Assistant
  📆 December 2019 - May 2021
  📍 University of Texas at Arlington - Arlington TX, USA
  Computer Vision Researcher at VLM (Vision-Learning-Mining) Lab • Supervisor: Dr Vassillis Athitsos • 05/2020 – 09/2021
  - Work involved data preprocessing from depth and color modalities, human pose estimation, activity detection and recognition
  - Collaborate with other researchers on activity recognition and cognitive assessment in children through motion capture and computer vision.
  - Master’s thesis: Structure Aware Human Pose Estimation using Adversarial Learning.
  Autonomous Vehicle Development at Sigma Lab • Supervisor: Dr Chen Kan • 12/2019 – 05/2020
  - Integrated ROS on Jetson TX2, an NVIDIA AI platform for a self-driving racecar.
  - Used LIDAR, Zed Stereo Camera data inputs and designed a CNN model for autonomous lane navigation.

• 💼 Graduate Teaching Assistant
  📆 January 2020 - August 2021
  📍 University of Texas at Arlington - Arlington TX, USA
  Courses: Operating Systems CSE 3320, Discrete Structures CSE 2315, Object Oriented Programming CSE 1325, and CSE1310 Introduction to Computers and Programming
  - Assisted faculty members with online instructional preparation, delivery, and assessment.
  - Evaluating and grading projects using a one-on-one virtual demo, through Microsoft Teams as well as leading two discussion sessions every week.

Technology & Tools

Programming Languages
Python Java JavaScript Modern C++ Shell MATLAB

Deep Learning Frameworks and ML Libraries
PyTorch TensorFlow 2.0 TensorFlow.JS Deeplearning4j scikit-learn OpenCV NumPy SciPy Pandas Matplotlib CUDA gglplot Tableau

Databases
MySQL AWS RDS AWS DynamoDB MongoDB

Web Technologies
HTML5 CSS3 React.JS Spring REST API XML JSON Redis

Cloud
AWS (EC2, S3, RDS, DynamoDB, AWS Lambda, ECS) MS Azure (Azure ML)

Data Analysis and Visualization
PySpark NumPy Pandas Matplotlib Tableau gglplot

Web Application Development
Spring Boot Docker REST API HTML5 CSS3 D3.JS React.JS Node.JS

Software Development
Unix Git Agile Development JUnit SCRUM PyCharm IntelliJ IDEA Visual Studio Eclipse Mathworks MATLAB

---

Contact

adv-pose-estimation's Issues

how can we test in real-time video/webcam and how to run it on android device

hi, could you please provide code for running it on realtime webcam and how to convert it into andriod device support format and run it on the andriod device.