Update README.md

daeunni · Feb 16, 2024 · 4c1c903 · 4c1c903
1 parent 35ec94f
commit 4c1c903
Showing 1 changed file with 20 additions and 2 deletions.
diff --git a/README.md b/README.md
@@ -1,7 +1,8 @@
 # BECoTTA: Input-dependent Online Blending of Experts for Continual Test-time Adaptation
-> The code will be updated soon. 
+[![Project Website](https://img.shields.io/badge/Project-Website-blue)](https://becotta-ctta.github.io/)  [![arXiv](https://img.shields.io/badge/arXiv-2402.08712-b31b1b.svg)](https://arxiv.org/pdf/2402.08712.pdf)   
 
-Daeun Lee*, Jaehong Yoon*, Sung Ju Hwang (* denotes equal contribution) 
+- Authors: [Daeun Lee*](https://daeunni.github.io/),  [Jaehong Yoon*](https://jaehong31.github.io/),  [Sung Ju Hwang](http://www.sungjuhwang.com/) (* denotes equal contribution)    
+> More details will be updated soon! 
 
 <img width="928" alt="image" src="https://github.com/daeunni/BECoTTA/assets/62705839/511417bf-bb8b-46c2-81ec-7ab9c208b92e">
 
@@ -10,3 +11,20 @@ Continual Test Time Adaptation (CTTA) is required to adapt efficiently to contin
 CTTA, forgetting-adaptation trade-offs and efficiency are still unexplored. Moreover, current CTTA scenarios assume only the disjoint situation, even though real-world domains are seamlessly changed. 
 To tackle these challenges, this paper proposes BECoTTA, an input-dependent yet efficient framework for CTTA. We propose Mixture-of-Domain Low-rank Experts (MoDE) that contains two core components: i) Domain-Adaptive Routing, which aids in selectively capturing the domain-adaptive knowledge with multiple domain routers, and (ii) Domain-Expert Synergy Loss to maximize the dependency between each domain and expert. We validate our method outperforms multiple CTTA scenarios including disjoint and gradual domain shits, while only requiring ∼98% fewer trainable parameters. We also provide analyses of our method, including the construction of experts, the effect of domain-adaptive experts, and visualizations. 
 
+## CTTA (Continual Test-time Adaptation) 
+```
+bash becotta.sh
+```
+
+
+## Setup 
+### [1] Environment
+- We follow [mmsegmentation code base](https://drive.qin.ee/api/raw/?path=/cv/cvpr2022/acdc-seg.tar.gz) provided by [CoTTA](https://github.com/qinenergy/cotta) authors. 
+
+
+### [2] Dataset
+- You can download **ACDC dataset** from [here](https://acdc.vision.ee.ethz.ch/download).
+
+### [3] Our warm-up model 
+- We provide our trained initialized model checkpoints. 
+