Sundeep Joshi

• Spearheading the development of cutting-edge algorithms for next-generation Intraoral scanner technology.
• Conducting comprehensive research in 3D computer vision to innovate and enhance algorithmic solutions.
• Leading the end-to-end development lifecycle, from initial implementation to deploying features in production environments.
• Collaborating with cross-functional teams to ensure seamless integration of new algorithms into existing systems.
• Continuously optimizing and refining algorithms to improve accuracy, efficiency, and overall performance.
• Staying updated with the latest advancements in computer vision and related fields to incorporate best practices and novel techniques.

Designed and implemented end-to-end computer vision pipelines.

• Significantly reduced operational costs by shifting over 80% of the inference load from GPU to CPU (Intel), maintaining similar latency and accuracy. This optimization led to a substantial increase in the conversion rate of potential customers to actual customers.
• Introduced new research areas within the team to reduce training time and resource requirements using sparsity-aware training techniques.
• Developed proof-of-concept projects for digital twins aimed at telecom resource planning, management, and downtime prediction.
• Worked on proof-of-concept for streaming 3D volumetric video in real-time.

• Lead team of Engineers to develop 3D reconstruction algorithms using multiple Depth and RGB sensors.
• Researched, prototyped, developed and optimized state of the art computer vision algorithms like voxel carving, colour map optimisation point cloud compression and Neural Radiance Fields to run on data centre grade servers and edge devices.
• Proposed and realized ideas to improve performance by using state of the art Deep learning research.
• Collaborated with international teams to deliver software that impacts virtual meeting and learning experiences for several users.
• Received Performance excellence award and best team award for FY 21.

• Worked and Lead development of Experiential learning project.
• Developed algorithms that do gaze detection, speech to text, face recognition and Direction of arrival estimation using 360 cameras and multiple microphones on edge devices.
• Collaborated with international teams to deliver product that impacts millions of students and teachers by improving their experience learning and teaching experience.

• Prototyped and developed Computer vision algorithms to automate 3D reconstruction of assets for virtual Try-on application
• Contributed to head pose estimation algorithm development by training good quality facial landmarks estimator
• Presented talks on specific project related problems currently under consideration, importance of projects and breakdowns of outcomes versus objectives.
• Worked side by side with colleagues and teams to solve critical problems.