Commitment Period: Jan 2026 onwards (Full-time Internship, min. 6 months)
Job Description
We are looking for passionate and motivated Computer Science students to participate in evaluating and optimizing various existing CNN-based computer vision applications targeting edge devices.
Responsibilities:
- Work with ML team to bring up various computer vision applications (Object Detection, Segmentation, Pose Estimation, Face Recognition, Hand Gesture Detection etc.) targeting edge devices using open-source CNN models and datasets.
- Perform quantitative and qualitative evaluations of deployed models on edge devices. Iterate and improve models based on performance metrics using Quantization-Aware Training (QAT).
- Conduct analytic studies to identify new applications, model and datasets from current State-of-the-Art and competition.
- Maintain clear and comprehensive scripts & documentation of the deployed models, dataset preparation processes, and optimization techniques used.
Requirements:
- Pursuing degree in Computer Science, Computer Engineering or Electrical Engineering.
- Good communication, collaboration and learning skills.
- Proficiency in Python and experience with deep learning frameworks PyTorch, TensorFlow, MXNet, ONNX.
- Experience with NumPy, MatPlotLib and other relevant Python packages for ML applications development.
- Research projects and contributions on open-source platforms like GitHub, HuggingFace etc. for CNN or Transformer based computer vision applications.
- Understanding of Edge AI concepts like PTQ, QAT, Tuning is desired.
- Commit at least 6 months full time and preferably 9 months.
