is not image segmentation?
I am so confused about this.
Thanks.
What is the method that locates an object(s) by labelling the pixels, where each similar object(s) is assigned to the same class?
Course Q&A
TensorFlow: Data and Deployment Specialization
Data Pipelines with TensorFlow Data Services
Yes, I refer to this,
but it told me that it was wrong about the question which is " what is the method that locates an object by labelling the pixels, where each similar object is assigned to the same class?".
Ah. I see. I think that in this case, it expected the answer saying “semantic segmentation.”
From Wikipedia:
Semantic segmentation is an approach detecting, for every pixel, belonging class of the object.[14] For example, when all people in a figure are segmented as one object and background as one object.
Instance segmentation is an approach that identifies, for every pixel, a belonging instance of the object. It detects each distinct object of interest in the image.[15] For example, when each person in a figure is segmented as an individual object.
Both are kinds of image segmentation, though.