Skip to content

Code and Dataset for ICCV 2019 paper. "Depth-induced Multi-scale Recurrent Attention Network for Saliency Detection".

License

Notifications You must be signed in to change notification settings

hongminglin08/DMRA_RGBD-SOD

 
 

Repository files navigation

DMRA_RGBD-SOD

Code repository for our paper entilted "Depth-induced Multi-scale Recurrent Attention Network for Saliency Detection" accepted at ICCV 2019 (poster).

Overall

avatar

The proposed Dataset

  • Dataset: DUTLF
  1. This dataset consists of DUTLF-MV, DUTLF-FS, DUTLF-Depth.
  2. The dataset will be expanded to 3000 about real scenes.
  3. We are working on it and will make it publicly available soon.
  • Dataset: DUTLF-Depth
  1. The dataset is part of DUTLF dataset captured by Lytro camera, and we selected a more accurate 1200 depth map pairs for more accurate RGB-D saliency detection.
  2. We create a large scale RGB-D dataset(DUTLF-Depth) with 1200 paired images containing more complex scenarios, such as multiple or transparent objects, similar foreground and background, complex background, low-intensity environment. This challenging dataset can contribute to comprehensively evaluating saliency models.

avatar

  • The dataset link can be found here. And we split the dataset including 800 training set and 400 test set.

DMRA Code

> Requirment

  • pytorch 0.3.0+
  • torchvision
  • PIL
  • numpy

> Usage

1. Clone the repo

git clone https://github.com/jiwei0921/DMRA.git
cd DMRA/

2. Train/Test

  • test
    Download related dataset link, and set the param '--phase' as "test" and '--param' as 'True' in demo.py. Meanwhile, you need to set dataset path and checkpoint name correctly.
python demo.py
  • train
    Our train-augment dataset link [ fetch code haxl ] / train-ori dataset, and set the param '--phase' as "train" and '--param' as 'True'(loading checkpoint) or 'False'(no loading checkpoint) in demo.py. Meanwhile, you need to set dataset path and checkpoint name correctly.
python demo.py

> Training info and pre-trained models for DMRA

To better understand, we retrain our network and record some detailed training details as well as corresponding pre-trained models.

Iterations Loss NJUD(F-measure) NJUD(MAE) NLPR(F-measure) NLPR(MAE) download link
100W 958 0.882 0.048 0.867 0.031 link
70W 2413 0.876 0.050 0.854 0.033 link
40W 3194 0.861 0.056 0.823 0.037 link
16W 8260 0.805 0.081 0.725 0.056 link
2W 33494 0.009 0.470 0.030 0.452 link
0W 45394 - - - - -
  • Tips: The results of the paper shall prevail. Because of the randomness of the training process, the results fluctuated slightly.

> Results

| DUTLF-Depth | | NJUD | | NLPR | | STEREO | | LFSD | | RGBD135 | | SSD |

  • Note: For evaluation, all results are implemented on this ready-to-use toolbox.

> Related RGB-D Saliency Datasets

All common RGB-D Saliency Datasets we collected are shared in ready-to-use manner.

  • The web link is here.

If you think this work is helpful, please cite

@InProceedings{Piao_2019_ICCV,       
   author = {Yongri {Piao} and Wei {Ji} and Jingjing {Li} and Miao {Zhang} and Huchuan {Lu}},   
   title = {Depth-induced Multi-scale Recurrent Attention Network for Saliency Detection},     
   booktitle = "ICCV",     
   year = {2019}     
}  

Contact Us

If you have any questions, please contact us ( [email protected] ).

About

Code and Dataset for ICCV 2019 paper. "Depth-induced Multi-scale Recurrent Attention Network for Saliency Detection".

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%