Recurrent Neural Networks for
Semantic Instance Segmentation
Amaia Salvador Jordi Torres Xavier Giró-i-Nieto Ferran MarquésManel BaradadMíriam Bellver
Motivation
2
Semantic Instance Segmentation
- Proposal-based solutions:
- Hundreds/Thousands of redundant predictions
- Post-processing needed (NMS)
- Holistic & class-agnostic methods
- Reduced set of predictions
- Separate network for semantics
Motivation
3
- Our solution:
- Recurrent model that sequentially predicts binary masks and
categorical labels for each object in an image.
- Learns to stop once all objects have been found.
- Does not need post-processing on its output.
Semantic Instance Segmentation
Model
4
Model
5
Pascal VOC
6
Average Precision at different IoU thresholds
Pascal VOC
7
Average Recall at different IoU thresholds (class-agnostic evaluation)
Pascal VOC
8
CVPPP Leaves Segmentation
9
Cityscapes
10
Object Sorting Patterns
11
Encoder Activations
Correlation with convolutional activations in the encoder before and after training
Encoder Activations
Activations at the end of the encoder before and after training

Recurrent Neural Networks for Semantic Instance Segmentation