Update bayesian segnet

Author: patrick-llgc

paper_notes/bayesian_segnet.md

@@ -2,16 +2,22 @@
 
 _June 2019_
 
-tl;dr: Extension of SegNet and output uncertainty map alongside the segmentation map.
+tl;dr: Estimate the variance of segmentation uncertainty with dropout inference samples. Use the mean for prediction. The idea is quite similar to TTA (test time augmentation).
 
 #### Overall impression
-Describe the overall impression of the paper. 
+The paper provides a practical way to evaluate the uncertainty (this is the epistemic uncertainty), at a cost at inference time.  Refer to [Bayesian DL](uncertainty_bdl.md) for integration with aleatoric uncertainty. 
 
 #### Key ideas
-- Summaries of the key ideas
+- Sampling with dropout performs better than weight averaging (normal dropout behavior during eval). Sampling dropout performs better than weight averaging after approximation with 6 samples. The performance saturates with 40 samples. 
+	- This comes at inference time cost, but is naively parallelizable.
+- The results also show that when the model predicts an incorrect label the model uncertainty is very high.
+- Class boundaries usually display high level of uncertainty.
+- Objects that are occluded or at a distance from the camera are are uncertain.
+- The uncertainty score is inversely proportional to occurrence and accuracy. The model is more confident about classes which are easier and occur more often.
+- The accuracy improves when we use a tighter threshold to filter out non-confident results. Uncertainty is an effective measure of accuracy.
 
 #### Technical details
-- Summary of technical details
+- **No need to use dropout layer after every layer.** Get the optimal architecture first by test placing dropout in different places. Then keep using dropout during inference (variational inference).
 
 #### Notes
 - [poster](https://alexgkendall.com/media/presentations/bmvc17_bayesian_segnet_poster.pdf)