cv
Basics
| Name | Hitesh Kumar |
| Label | Deep Learning and Robotics Researcher |
| hitesh.kumar61@yahoo.com | |
| Url | https://hitesh.one |
| Summary | In the long term, I envision being the best at what I do. I believe an MS and a dynamic research culture will help me get there.. |
Work
-
2022.08 - Present Delhi, India
Deep Learning Engineer - Autonomous R&D
Euler Motors
Developing ADAS and Night Vision Assist for electric cargo vehicles.
- Built real-time ADAS using C++ for Nvidia Jetson Nano and Quectel SC600
- Implemented forward collision warning with over 80% precision
- Developed ultralight classification model for traffic lights with 95% accuracy
- Optimized object detection and segmentation models, reducing size by 60%
- Collaborated on Night Vision Assist system using IR camera
-
2021.07 - 2022.08 Gurugram, India
Deep Learning Engineer - R&D
SynergyLabs
Developed AI solutions for smart city initiatives.
- Trained classification & object detection models, improving performance by 14%
- Implemented automatic number plate detection for urban traffic monitoring
- Developed Vehicle Detection System with less than 6% speed error
-
2021.02 - 2021.07 Gurugram, India
Deep Learning Intern - R&D
SynergyLabs
- Developed 'Automatic Traffic Counter System' deployed across 300+ locations
- Optimized MobilenetV2 model, reducing size to 2MB
- Created product configuration UI with PyQT, reducing manual effort by 3x
Education
-
2016.01 - 2020.01 Delhi, India
Publications
-
2021.01.01 Estimation Of Surface Roughness in turning operations using Multivariate Polynomial Regression
Advances in Industrial and Production Engineering
-
2020.01.01 Analytical and Computational Modelling of Go-Kart Powertrains
International Journal of Mechanical and Production Engineering Research and Development (IJMPERD)
Skills
| Programming Languages | |
| C++ | |
| Python | |
| Golang |
| Frameworks & Libraries | |
| OpenCV | |
| PyTorch | |
| TensorFlow | |
| CUDA | |
| NumPy | |
| Matplotlib | |
| FFMPEG |
| Tools & Platforms | |
| Linux | |
| Docker | |
| Google Cloud | |
| ROS | |
| CMake | |
| Git |
| Hardware & Sensors | |
| Nvidia Jetson | |
| Raspberry Pi | |
| Arduino | |
| 2D/4D Radar | |
| Lidar | |
| Intel RealSense | |
| IR Sensor | |
| Ultrasonic Sensor |
Languages
| English | |
| Fluent |
| Hindi | |
| Native |
| Korean | |
| Basic |
Interests
| Research Areas | |
| Computer Vision | |
| Autonomous Systems | |
| Deep Learning | |
| Perception | |
| Vision Language Models |
Projects
-
-
-
ReID Tracking using YOLOv8
Integrated YOLOv8 with ByteTrack and StrongSORT for re-identification in object tracking.
-
Light Object Detection
Developing a package for lightweight object detection models for edge device inference.
-
Self-driving Vision Stack
Implemented a primary self-driving vision stack including lane detection, car detection, traffic light detection, and pedestrian detection.
Volunteer
-
2018.01 - 2019.01 Software Head (Autonomous)
InfernoDTU (Project Mars Rover Prototype)
Led autonomous tech team in developing Mars Rover Prototype's autonomous features for competitive events.
- Utilized ROS in C++/Python for remote rover navigation
- Implemented computer vision techniques for obstacle detection and collision avoidance
-
2016.01 - 2018.01 Core Member
Society of Robotics DTU
Organized tech events and mentored juniors in computer vision fundamentals.