Phoenix - Your AI AssistantImage Segmentation Model —Semantic & Instance Segmentation Build
Delivery in
5 days
- Views 5
Amount of days required to complete work for this Offer as set by the freelancer.
Rating of the Offer as calculated from other buyers' reviews.
Average time for the freelancer to first reply on the workstream after purchase or contact on this Offer.
What you get with this Offer
I will train a custom image segmentation model for your use case — semantic segmentation for pixel-level class labelling (which pixels belong to which category) or instance segmentation for individual object mask generation — using SegFormer, DeepLab, or SAM fine-tuning depending on your requirements. Segmentation provides the most detailed spatial understanding of an image and is the right approach for applications where knowing an object's precise shape and extent matters — not just its presence (classification) or location (detection), but the exact pixel boundaries of each object or region in the scene.
The model covers mask annotation validation, backbone and head architecture selection, training with IoU and Dice loss, mean IoU evaluation per class, visualisation of predicted versus ground truth masks, and an inference API returning mask outputs.
The model covers mask annotation validation, backbone and head architecture selection, training with IoU and Dice loss, mean IoU evaluation per class, visualisation of predicted versus ground truth masks, and an inference API returning mask outputs.
What the Freelancer needs to start the work
Please share your annotated dataset (COCO format preferred for instance, PNG masks for semantic), your segmentation task description, your deployment target, and your inference latency requirements.