Saliency-Ranking

KO | EN

gitlite — search

#javascript #python #hacktoberfest #react #ai #typescript #llm #go #golang #android #machine-learning #rust #deep-learning #linux

Saliency-Ranking

★ 16 Open GitHub ↗

Code release for the TPAMI 2021 paper "Instance-Level Relative Saliency Ranking with Graph Reasoning" by Nian Liu, Long Li, Wangbo Zhao, Junwei Han, and Ling Shao.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Saliency-Ranking
Code release for the TPAMI 2021 paper "Instance-Level Relative Saliency Ranking with Graph Reasoning" by Nian Liu, Long Li, Wangbo Zhao, Junwei Han, and Ling Shao.

![avatar](https://raw.githubusercontent.com/dragonlee258079/Saliency-Ranking/main/structure.png)

&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;

## Installation
See [INSTALL.md](INSTALL.md).

&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;

## Data Preparation
Download the datatset from [Baidu Driver](https://pan.baidu.com/s/1p8XVoBllw-re-fnynPodZA) (rx96) or [Google Driver](https://drive.google.com/file/d/1fp-LdlmTdoWsEK1jxP6VMjY-mJ1PhoQR/view?usp=sharing) and unzip them to './dataset'. Then the structure of the './dataset' folder will show as following:

````
-- dataset
   |-- Annotations
   |   |-- | train.pkl
   |   |-- | test.pkl
   |-- Images
   |   |-- train
   |   |-- |-- | rgb
   |   |-- |-- |-- | COCO_train2014_000000000110.jpg ...
   |   |-- |-- | gt
   |   |-- |-- |-- | COCO_train2014_000000000110.png ...
   |   |-- test
   |   |-- |-- | rgb
   |   |-- |-- |-- | COCO_val2014_000000000192.jpg ...
   |   |-- |-- | gt
   |   |-- |-- |-- | COCO_val2014_000000000192.png ...
````

&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;

## Training model
1. Download the pretrained model (our modified Mask R-CNN model for salient instance segmentation) from [Baidu Driver](https://pan.baidu.com/s/1O_3vCeruILrLcVkqnZauWQ)(spq9) or [Google Driver](https://drive.google.com/file/d/1gyH89OVRzgvtVqyXr61y3qnGbEvwuosl/view?usp=sharing) and put it into `./model` folder.
2. Run `python ./tool/plain_train_net.py`. 
3. The trained model will be saved in `./output` folder. Additionally, the evaluaion results produced during training process will be saved    in `./output/SA_SOR.txt` and            `./output/MAE.txt`. 
## Testing model
1. Download our trained model from [Baidu Driver](https://pan.baidu.com/s/1otL_RvGaS1f91TsqvLg2ng)(fhz7) or [Google Driver](https://drive.google.com/file/d/1zBylV2C2BV_Is6yNQkNHCn4XIz27Xpr0/view?usp=sharing). Rename it as 'final_model.pth' and put it into `./model` folder.
2. Run `python ./tool/plain_test_net.py`.
3. The prediction images will be saved in `./prediction`.  And the metric score, **SA_SOR** and MAE, will be printed.

&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;

## Result
![alt_text](https://raw.githubusercontent.com/dragonlee258079/Saliency-Ranking/main/qualitative.png)
The prediction results of our dataset can be download from [prediction](https://pan.baidu.com/s/1ysWXDwHhDU_MQZL8s8xt9Q) (k1jr).

## Metric
We propose a new evaluation metirc for this task, which comprehensively considers salient instance detection, segmentation, and ranking performance. See [SA-SOR](./metric).

We test the SOR metric of our model by first predicting the saliency maps via the test.py in our code and then feeding these saliency maps to the official SOR code (a matlab code can be download from [SOR code](https://pan.baidu.com/s/19onIw5tztUOG1hpHWNHsHw)(dems)).

Saliency-Ranking

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README