Skip to content

Commit

Permalink
readme update
Browse files Browse the repository at this point in the history
  • Loading branch information
plemeri committed Nov 15, 2022
1 parent aaa67ec commit eb0e14d
Showing 1 changed file with 24 additions and 24 deletions.
48 changes: 24 additions & 24 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -48,11 +48,11 @@ To appear in the 16th Asian Conference on Computer Vision (ACCV2022)
5. [Getting Started](#flight_departure-getting-started)
6. [Model Zoo](#giraffe-model-zoo)
7. [Results](#100-results)
* [Quantitative Results](#quantitative-results)
* [Qualitative Results](#qualitative-results)
* [Quantitative Results](#quantitative-results)
* [Qualitative Results](#qualitative-results)
8. [Citation](#trophy-citation)
9. [Acknowledgement](#mortar_board-acknowledgement)
* [Special Thanks to](#tada-special-thanks-to)
* [Special Thanks to](#tada-special-thanks-to)
10. [References](#paperclip-references)

## :newspaper: News
Expand Down Expand Up @@ -117,33 +117,33 @@ Please refer to [model_zoo.md](./docs/model_zoo.md) for downloading pre-trained

### :new: [Added in 2022.10.20] Quantitative Results on DIS5K Dataset [[Log file](https://postechackr-my.sharepoint.com/:t:/g/personal/taehoon1018_postech_ac_kr/EeczZ1XEboZKhxqif9m1VwsBhMc--dLYqlZ_5TicEXr2ZA?e=aCFXhp)]

<details><summary>Click to see the results</summary>
<p>
<details><summary>Click to see the results</summary>
<p>

* Trained with LR scale (384 X 384) [[Config](./configs/extra_dataset/InSPyReNet_SwinB_DIS5K_LR.yaml)]
* Trained with LR scale (384 X 384) [[Config](./configs/extra_dataset/InSPyReNet_SwinB_DIS5K_LR.yaml)]

Dataset | Sm | mae | adpEm | maxEm | avgEm | adpFm | maxFm | avgFm | wFm | mBA | HCE<sup>*</sup> |
:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |
DIS-VD | 0.8868 | 0.0427 | 0.9145 | 0.9352 | 0.9217 | 0.8295 | 0.8760 | 0.8523 | 0.8259 | 0.7654 | 905 |
DIS-TE1 | 0.8618 | 0.0447 | 0.8679 | 0.9071 | 0.8952 | 0.7556 | 0.8341 | 0.8083 | 0.7771 | 0.7453 | 148 |
DIS-TE2 | 0.8934 | 0.0383 | 0.9131 | 0.9356 | 0.9253 | 0.8281 | 0.8811 | 0.8599 | 0.8339 | 0.7587 | 316 |
DIS-TE3 | 0.9019 | 0.0381 | 0.9278 | 0.9496 | 0.9376 | 0.8529 | 0.9038 | 0.8802 | 0.8558 | 0.7741 | 582 |
DIS-TE4 | 0.8913 | 0.0461 | 0.9316 | 0.9433 | 0.9255 | 0.8545 | 0.8915 | 0.8655 | 0.8395 | 0.7789 | 2243 |
Dataset | Sm | mae | adpEm | maxEm | avgEm | adpFm | maxFm | avgFm | wFm | mBA | HCE<sup>*</sup> |
:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |
DIS-VD | 0.8868 | 0.0427 | 0.9145 | 0.9352 | 0.9217 | 0.8295 | 0.8760 | 0.8523 | 0.8259 | 0.7654 | 905 |
DIS-TE1 | 0.8618 | 0.0447 | 0.8679 | 0.9071 | 0.8952 | 0.7556 | 0.8341 | 0.8083 | 0.7771 | 0.7453 | 148 |
DIS-TE2 | 0.8934 | 0.0383 | 0.9131 | 0.9356 | 0.9253 | 0.8281 | 0.8811 | 0.8599 | 0.8339 | 0.7587 | 316 |
DIS-TE3 | 0.9019 | 0.0381 | 0.9278 | 0.9496 | 0.9376 | 0.8529 | 0.9038 | 0.8802 | 0.8558 | 0.7741 | 582 |
DIS-TE4 | 0.8913 | 0.0461 | 0.9316 | 0.9433 | 0.9255 | 0.8545 | 0.8915 | 0.8655 | 0.8395 | 0.7789 | 2243 |

* Trained with HR scale (1024 X 1024) [[Config](./configs/extra_dataset/InSPyReNet_SwinB_DIS5K.yaml)]
* Trained with HR scale (1024 X 1024) [[Config](./configs/extra_dataset/InSPyReNet_SwinB_DIS5K.yaml)]

Dataset | Sm | mae | adpEm | maxEm | avgEm | adpFm | maxFm | avgFm | wFm | mBA | HCE<sup>*</sup> |
:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |
DIS-VD | 0.8997 | 0.0416 | 0.9141 | 0.9398 | 0.9216 | 0.8364 | 0.8893 | 0.8603 | 0.8336 | 0.8003 | 904 |
DIS-TE1 | 0.8733 | 0.0432 | 0.8743 | 0.9070 | 0.8940 | 0.7771 | 0.8453 | 0.8195 | 0.7882 | 0.7966 | 110 |
DIS-TE2 | 0.9053 | 0.0364 | 0.9154 | 0.9410 | 0.9279 | 0.8439 | 0.8941 | 0.8703 | 0.8462 | 0.8034 | 255 |
DIS-TE3 | 0.9179 | 0.0341 | 0.9403 | 0.9586 | 0.9427 | 0.8752 | 0.9191 | 0.8923 | 0.8707 | 0.8079 | 522 |
DIS-TE4 | 0.9052 | 0.0424 | 0.9361 | 0.9535 | 0.9282 | 0.8598 | 0.9051 | 0.8700 | 0.8479 | 0.7994 | 2336 |
Dataset | Sm | mae | adpEm | maxEm | avgEm | adpFm | maxFm | avgFm | wFm | mBA | HCE<sup>*</sup> |
:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |:- |
DIS-VD | 0.8997 | 0.0416 | 0.9141 | 0.9398 | 0.9216 | 0.8364 | 0.8893 | 0.8603 | 0.8336 | 0.8003 | 904 |
DIS-TE1 | 0.8733 | 0.0432 | 0.8743 | 0.9070 | 0.8940 | 0.7771 | 0.8453 | 0.8195 | 0.7882 | 0.7966 | 110 |
DIS-TE2 | 0.9053 | 0.0364 | 0.9154 | 0.9410 | 0.9279 | 0.8439 | 0.8941 | 0.8703 | 0.8462 | 0.8034 | 255 |
DIS-TE3 | 0.9179 | 0.0341 | 0.9403 | 0.9586 | 0.9427 | 0.8752 | 0.9191 | 0.8923 | 0.8707 | 0.8079 | 522 |
DIS-TE4 | 0.9052 | 0.0424 | 0.9361 | 0.9535 | 0.9282 | 0.8598 | 0.9051 | 0.8700 | 0.8479 | 0.7994 | 2336 |

* *: HCE here is relax Human Correction Error which is proposed in DIS([project page](https://xuebinqin.github.io/dis/index.html) | [paper](https://arxiv.org/pdf/2203.03041.pdf)) and you can compute yourself from their [github](https://github.com/xuebinqin/DIS) repository.
* *: HCE here is relax Human Correction Error which is proposed in DIS([project page](https://xuebinqin.github.io/dis/index.html) | [paper](https://arxiv.org/pdf/2203.03041.pdf)) and you can compute yourself from their [github](https://github.com/xuebinqin/DIS) repository.

</p>
</details>
</p>
</details>

### Qualitative Results

Expand Down

0 comments on commit eb0e14d

Please sign in to comment.