From 1a4ce295405e727e3e85789c0b69955c802ede12 Mon Sep 17 00:00:00 2001 From: Himalaya Jain Date: Tue, 11 Jun 2019 17:29:20 +0200 Subject: [PATCH] Update README.md --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index d7eed97..a3314b5 100644 --- a/README.md +++ b/README.md @@ -5,7 +5,7 @@ Code will be available soon. ![](figures/teaser.jpg) [ADVENT: Adversarial Entropy Minimization for Domain Adaptation in Semantic Segmentation](https://arxiv.org/abs/1811.12833) - [Tuan-Hung Vu](https://tuanhungvu.github.io/), [Himalaya Jain](https://scholar.google.fr/citations?user=Xl7SNlsAAAAJ), [Maxime Bucher](https://maximebucher.github.io/), [Matthieu Cord](http://webia.lip6.fr/~cord/), [Patrick Pérez](https://ptrckprz.github.io/) + [Tuan-Hung Vu](https://tuanhungvu.github.io/), [Himalaya Jain](https://himalayajain.github.io/), [Maxime Bucher](https://maximebucher.github.io/), [Matthieu Cord](http://webia.lip6.fr/~cord/), [Patrick Pérez](https://ptrckprz.github.io/) valeo.ai, France IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019 (**Oral**) @@ -24,4 +24,4 @@ If you find this code useful for your research, please cite our [paper](https:// Semantic segmentation is a key problem for many computer vision tasks. While approaches based on convolutional neural networks constantly break new records on different benchmarks, generalizing well to diverse testing environments remains a major challenge. In numerous real world applications, there is indeed a large gap between data distributions in train and test domains, which results in severe performance loss at run-time. In this work, we address the task of unsupervised domain adaptation in semantic segmentation with losses based on the entropy of the pixel-wise predictions. To this end, we propose two novel, complementary methods using (i) an entropy loss and (ii) an adversarial loss respectively. We demonstrate state-of-the-art performance in semantic segmentation on two challenging *synthetic-2-real* set-ups and show that the approach can also be used for detection. ## Demo -[![](http://img.youtube.com/vi/Ihmz0yEqrq0/0.jpg)](http://www.youtube.com/watch?v=Ihmz0yEqrq0 "") \ No newline at end of file +[![](http://img.youtube.com/vi/Ihmz0yEqrq0/0.jpg)](http://www.youtube.com/watch?v=Ihmz0yEqrq0 "")