(Deep Learning | Computer Vision | Software) Engineer
Experienced and results-driven Machine Learning & Software Engineer with a robust background spanning 3 years, navigating diverse technologies and industries encompassing both service and product domains. My passion lies in the intricate realm of Computer Vision, where I possess a deep understanding of camera imaging principles and fundamentals.
My expertise extends across a spectrum of advanced techniques, including Multi-Task Learning, Self-Supervised Learning, Continual Learning, Kalman Filters, and Object Tracking. I specialize in Image Segmentation, Optical Flow, Stereo Vision, Depth Estimation, and Sensor Fusion involving both Camera and LiDAR technologies.
Adept at crafting and implementing custom neural networks and distributed training, my skills shine in optimizing models for reduced memory consumption, compact size, and low inference latency. I bring to the table proficiency in cutting-edge techniques such as Knowledge Distillation, Pruning, and Quantization. My deployment capabilities encompass various runtime accelerators, including ONNXRuntime, Intel OpenVINO, Nvidia TensorRT, and the Nvidia Triton Inference Server for seamless serving of single or ensemble of models.
Beyond Computer Vision, I excel in backend application development using FastAPI, ensuring scalability and performance. My proficiency extends to deploying applications seamlessly using cloud-managed CI/CD pipelines, Docker, and Kubernetes.
I am passionate about pushing the boundaries of what's possible in the world of Machine Learning and Computer Vision, continually seeking innovative solutions to complex challenges.
Extending a Multi-Task Network that does Monocular Depth Estimation and Semantic Segmentation to also perform Object Detection
Self Supervised Learning for Monocular Depth Estimation
Compressing a Monocular Depth Estimation Regressor Network
Real-time Multi-Task Learning for Indoor and Outdoor scenes using Aysmmetric Annotations
