4 papers
Soft Self-labeling and Potts Relaxations for Weakly-Supervised Segmentation
Zhongwen Zhang, Yuri Boykov
We consider weakly supervised segmentation where only a fraction of pixels have ground truth labels (scribbles) and focus on a self-labeling approach optimizing relaxations of the…
Approximate Size Targets Are Sufficient for Accurate Semantic Segmentation
Xingye Fan, Zhongwen, Zhang +1
This paper demonstrates a surprising result for segmentation with image-level targets: extending binary class tags to approximate relative object-size distributions allows off-the-…
Collision Cross-entropy for Soft Class Labels and Deep Clustering
Zhongwen Zhang, Yuri Boykov
We propose "collision cross-entropy" as a robust alternative to Shannon's cross-entropy (CE) loss when class labels are represented by soft categorical distributions y. In general,…
Discriminative Entropy Clustering and its Relation to K-means and SVM
Zhongwen Zhang, Yuri Boykov
Maximization of mutual information between the model's input and output is formally related to "decisiveness" and "fairness" of the softmax predictions, motivating these unsupervis…