0

I have 3 images of differents objets : a smartphone, a shirt and a packet of pasta. I want to perform recognition of each object on any images containing one of these objects. For example, if we have the same phone in a picture, i want to be able to see the phone with a bounded box drawn in this picture. If the phone is different, nothing should be drawn.

I first tried to perform object recognition using neural network like Mask R-CNN with python and tensorflow. But i realized that i haven't a huge training dataset, only my 3 images. Neural network algorithms seem to be adapted to recognize concept like dog, smartphone, landscape but not a particular dog, a specific smartphone or a specific landscape.

To get to the point, if i have in input any picture that contain the same smartphone, the same shirt or the same packet of pasta, i want the program to detect that.

What algorithms are best suited to perform this recognition ?

samourayonly
  • 181
  • 1
  • 1
  • 6

1 Answers1

0

Try using the COCO dataset. Since the COCO weights have already been trained on thousands of items and images, you should just be able to run the splash feature to help detection with Mask RCNN.

enter image description here

Worst case scenario, if you want to train your own dataset, just find a lot of photos online relating to the objects you want to detect, annotate them, then train.

Aaron Jones
  • 1,140
  • 1
  • 9
  • 26
  • Thank you for your answer @Aaron Jones . But as i said in the question, i want to detect specific objects instances. I don't want to recognize generic objects like persons, airplanes, bottles for example, but recognize Aaron(a specific person), a Boeing 787(a specific airplane), a coke bottle(a specific type of bottle). Train a dataset over thousand of pictures is adapted to recognize generic objets i think. – samourayonly Apr 15 '20 at 18:22
  • @samourayonly You will most likely have to collect a bunch of images of the spacific object that you want, then train a model based on that, since you want the dataset to be very specific, and you have a limited beginning images. – Aaron Jones Apr 15 '20 at 18:28