Our objective is to collect a COCO-like object detection dataset containing 15 daily life objects: wall switches, wall sockets, doors, shoes, eyeglasses, domestic trash cans, paper sheets (A4), headphones, wireless headphones, laptop or phone charger, pen/pencils, lighters, wallets, keys, and cards.
This image dataset would contain 500 images per class (7 500 in total). The images would be daily scenes of these objects in various credible indoor contexts. If possible, we would like the photos to be taken with smartphones.
- For nonmoveable objects (switches, outlets, etc…), you would need to take pictures of different types, angles, lighting, etc…
- for moveable objects, you could move the object into different locations, lighting, background, ground, etc…
- Collection completed before December
- Collection country location doesn’t matter
- Collection needs to happen indoor
- Goal is the diversity of location of objects
Please make 15 folders for 15 objects
No annotation is required but you can submit a quote with annotation.
17 pekerja bebas membida secara purata $442 untuk pekerjaan ini
Hi, I am an expert in Graphic design and photography. Here is a link to my portfolio https://www.freelancer.com/u/machexport Lets discuss and move ahead. Thank you