# SSH_alignment

**Repository Path**: deepinx/SSH_alignment

## Basic Information

- **Project Name**: SSH_alignment
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2019-02-08
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

## Enhanced SSH with landmark localization

Add landmark localization using ONet of MTCNN to original SSH method

Widerface result, by pyramid testing.

| Impelmentation     | Easy-Set | Medium-Set | Hard-Set |
| ------------------ | -------- | ---------- | -------- |
| Original Caffe SSH | 0.93123  | 0.92106    | 0.84582  |
| Our Model          | 0.93394  | 0.92187    | 0.83682  |

Pre-trained models on [baiducloud](https://pan.baidu.com/s/1sghM7w1nN3j8-UHfBHo6rA) 

Download the pre-trained models and place it in *`./model/`*

You can use `python test.py` to test this alignment method.

## Test Environment

-   Python 2.7 
-   Ubuntu 18.04
-   Mxnet-cu90 (=1.3.0)

## Alignment Result
![detection result](https://raw.githubusercontent.com/deepinx/SSH_alignment/master/sample-images/detection_result.png)

## Contact

    xcliu1893[at]gmail.com