Train an existing computer vision machine learning object detection model for 6 new classes/objects
Budget: $250 – $750 USD
The project is simple yet can be tricky.
1. I will supply you with 50+ images for each of 4 objects from various angles
2. You should gather the images for an 2 extra comparable objects in your vicinity (choice of objects is a subject of an approval by me) as a secondary test and to help you debug
3. you need to select any of the existing machine learning models and re-train it so that the objects can be reliably detected
4. put together a super simple mobile app which detects the newly added objects in live camera view
5. write a simple 1 page document summarizing your approach or jump on a recorded video call to capture your process
if you achieve in 4/6 objects recognized (in daytime conditions) and deliver the report of your approach, you get paid the full contract and might also earn an ongoing consulting job. If your solution works 3/6, you get 50% of the pay. If it works at a lesser reliability, no pay.
sample of images attached
1. I will supply you with 50+ images for each of 4 objects from various angles
2. You should gather the images for an 2 extra comparable objects in your vicinity (choice of objects is a subject of an approval by me) as a secondary test and to help you debug
3. you need to select any of the existing machine learning models and re-train it so that the objects can be reliably detected
4. put together a super simple mobile app which detects the newly added objects in live camera view
5. write a simple 1 page document summarizing your approach or jump on a recorded video call to capture your process
if you achieve in 4/6 objects recognized (in daytime conditions) and deliver the report of your approach, you get paid the full contract and might also earn an ongoing consulting job. If your solution works 3/6, you get 50% of the pay. If it works at a lesser reliability, no pay.
sample of images attached