Revisiting OCR: Object Contextual Representations

Looking Back to Look Forward

OCR[1] was mainly motivated by the fact that the label of a pixel is the category of the object that the pixel belongs to.

So OCR augments pixels with Object Representations and at the same time learning the weights between the pixels and the object region representations.

Logically, OCR augments pixels in three steps

1. Construct soft region generator by learning from ground truth segmentation labels
2. Compute object region representations by considering all pixels within the soft region.
3. Calculate the relationship between current pixel and all object regions, and augment the representations of current pixel.

References

[1] Yuan, Yuhui, Xilin Chen, and Jingdong Wang. “Object-contextual representations for semantic segmentation.” Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part VI 16. Springer International Publishing, 2020.