Were hiring someone with real computer vision experience detection, segmentation, tracking, pose who can run experiments end to end and ship models into working systems.

Full-time or long-term contract (3–6 months).

What you’ll work on
Image and video understanding
Object detection, segmentation, keypoint detection, pose estimation, and tracking
Temporal analysis of actions and multi-step processes
Multimodal learning using visual, sensor, and time-series data
Failure detection, confidence estimation, and automated verification
Data collection, annotation, training, and evaluation pipelines
Integrating trained models into larger software systems

You’ll review research, set baselines, compare model families, dig into failures, and recommend the practical path.

What we’re looking for
Strong Python and PyTorch
Solid up-to-date computer vision and deep-learning architecture chops
Hands-on work with image or video datasets
Experience in detection, segmentation,



tracking, pose estimation, or action recognition
Designing controlled experiments and meaningful metrics
Linux and GPU-based training
Reading papers and turning them into working experiments
Comfort with data collection, labelling, train/val, and model integration

Valuable to have
Detection & segmentation: YOLO, RT-DETR, Mask R-CNN
Pose & tracking: ViTPose, Keypoint R-CNN, ByteTrack
Video / temporal: VideoMAE, Video Swin, MS-TCN
Learning-based control: Behaviour Cloning, ACT, Diffusion Policy
Multimodal learning, synthetic data, domain adaptation, uncertainty, or robotics apps

Who fits

Someone who owns an experimentation track — problem, data, models, comparison, explicit recommendation. We want help deciding what to build, what to test, and how to prove it works.

📌 Machine Learning & Computer Vision Engineer Pune
🏢 InitX
📍 Pune