Understanding a black-box model is a major problem in domains that relies on model predictions in critical tasks. If solved, can help to evaluate the trustworthiness of a model. This thesis proposes a user-centric approach to black-box interpretability. It addresses the problem in semantic segmentation setting with an example of humanitarian remote sensing application for building detection. The question that drives this work was, Can existing methods for explaining black-box classifiers be used for a deep learning semantic segmentation model? We approached this problem with exploratory qualitative research involving a case study and human evaluation. The study showed that it is possible to explain a segmentation model with adapted methods ...
Deep learning (DL) is currently the dominant approach to image classification and segmentation, but ...
The state-of-the-art object detection and image classification methods can perform impressively on m...
International audienceThis paper explores different aspects of semantic segmentation of remote sensi...
Understanding a black-box model is a major problem in domains that relies on model predictions in cr...
Understanding a black-box model is a major problem in domains that relies on model predictions in cr...
National audienceIn this paper, a novel method to tackle semantic segmentation of very high resoluti...
International audienceDeep learning architectures have received much attention in recent years demon...
Deep learning architectures have received much attention in recent years demonstrating state-of-the-...
International audienceDeep learning architectures have received much attention in recent years demon...
International audienceIn this paper, a novel method to tackle semantic segmentation of very high res...
International audienceDeep learning architectures have received much attention in recent years demon...
International audienceDeep learning architectures have received much attention in recent years demon...
National audienceIn this paper, a novel method to tackle semantic segmentation of very high resoluti...
Semantic segmentation consists of the generation of a categorical map, given an image in which each ...
This is the final version. Available from SPIE via the DOI in this recordSemantic segmentation is on...
Deep learning (DL) is currently the dominant approach to image classification and segmentation, but ...
The state-of-the-art object detection and image classification methods can perform impressively on m...
International audienceThis paper explores different aspects of semantic segmentation of remote sensi...
Understanding a black-box model is a major problem in domains that relies on model predictions in cr...
Understanding a black-box model is a major problem in domains that relies on model predictions in cr...
National audienceIn this paper, a novel method to tackle semantic segmentation of very high resoluti...
International audienceDeep learning architectures have received much attention in recent years demon...
Deep learning architectures have received much attention in recent years demonstrating state-of-the-...
International audienceDeep learning architectures have received much attention in recent years demon...
International audienceIn this paper, a novel method to tackle semantic segmentation of very high res...
International audienceDeep learning architectures have received much attention in recent years demon...
International audienceDeep learning architectures have received much attention in recent years demon...
National audienceIn this paper, a novel method to tackle semantic segmentation of very high resoluti...
Semantic segmentation consists of the generation of a categorical map, given an image in which each ...
This is the final version. Available from SPIE via the DOI in this recordSemantic segmentation is on...
Deep learning (DL) is currently the dominant approach to image classification and segmentation, but ...
The state-of-the-art object detection and image classification methods can perform impressively on m...
International audienceThis paper explores different aspects of semantic segmentation of remote sensi...