Computer Vision Engineer

The Role 

We are seeking a highly motivated Computer Vision Engineer to work on advanced problems in 3D vision and multimodal AI within the Multimodal Intelligence and Real Applications (MIRA) Lab. The role focuses on designing and deploying robust pipelines for multi-view capture, multimodal synchronization, and multimodal data fusion, enabling large-scale, high-quality data generation and AI model development.

 Responsibilities:

• o Design and develop algorithms for 3D vision and multi-view geometry, including camera calibration, triangulation, 3D reconstruction, and motion estimation.
• o Build and optimize pipelines for multimodal synchronization, ensuring precise temporal alignment across video, audio, and sensor streams.
• o Implement multimodal fusion techniques (early/late fusion, feature-level fusion) for tasks such as human behavior understanding and scene analysis.
• o Develop scalable data pipelines for large-scale multimodal datasets, including preprocessing, annotation workflows, and quality assurance.
• o Train and deploy deep learning models for computer vision tasks such as detection, tracking, segmentation, and spatiotemporal modeling.
• o Work with multi-sensor systems (multi-view cameras, motion capture systems, EEG/audio streams) and ensure seamless integration between hardware and software components.
• o Optimize performance across the stack using GPU acceleration, parallel processing, and efficient memory management.
• o Translate research prototypes into robust, maintainable, and production-ready pipelines, ensuring reliability and scalability.

Essential Qualification:

Bachelors in Computer Science, Electronics, Electrical Engineering, or a related field. Computer Vision Engineer 3-6 years Masters degree in Computer Science, Electronics, Electrical Engineering, or a related field. 1-2 Year

Desirable Experience:

o Strong foundation in 3D computer vision and geometric methods, including multi-view geometry, SfM, SLAM, and camera calibration.

o Extensive hands-on experience with deep learning for vision, including spatiotemporal modeling and video understanding. o Proficiency in Python with strong experience in frameworks such as PyTorch/TensorFlow and libraries like OpenCV.

• o Experience with multi-sensor systems and multimodal data, including synchronization, calibration, and fusion across modalities.
• o Strong experience with GPU computing (CUDA), high-performance pipelines, and large-scale data processing.
• o Familiarity with motion capture systems, signal processing (audio/EEG), or real-time systems is highly desirable.

Remuneration: CTC up to INR 9,28,800 per annum, inclusive of applicable annual benefits.