K-Nearest-Neighbors-Hashing

KO | EN

gitlite — search

#javascript #python #hacktoberfest #ai #typescript #machine-learning #docker #nodejs #java #cli #kubernetes #pytorch #mcp #open-source

K-Nearest-Neighbors-Hashing

(★ 27)

Matlab implementation of "K-Nearest Neighbors Hashing" (CVPR2019)

K-Nearest-Neighbors-Hashing 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# K-Nearest-Neighbors-Hashing
Matlab implementation of "K-Nearest Neighbors Hashing".

### Dependencies :
* **MatLab >= R2016b** (GPU support knnsearch)
* GPU memory >= 10G
* MatLab <= R2016a (CPU only knnsearch)

### How to use ?
1. Download datasets and unzip to `./datasets`
2. Edit `SETTING` part in `demo.m`:
```matlab
%% SETTING
num_bits = 16;
dataset = 'cifar10'; % cifar10, mnist, labelme, places
feature_type = 'vggfc7'; % gist, vggfc7, uint (MINIST only), alexnet (Places205 only)

gpuDevice_ID = 1; % if CPU-only / you don't want to use GPU, just set it as -1

ITER = 10; % run ITER times
K = 20; % 200 for Places205
%%
```
For Places205, we recommend downloading the precomputed KNN index matrix from [GoogleDrive](https://drive.google.com/open?id=1RsO5XyinqgPlOgD7Z-Ow50_stYwd7-37) or [Baidu Pan](https://pan.baidu.com/s/1YoeV7d1dX233yCuwNEf2dA) (582l), then move it to `./model`. It could be time consuming to generate the results using CPU.

3. run `demo.m`.

### Demo results :
mAP | Cifar10-Gist | MNIST-Gist | LabelMe-Gist | Cifar10-VGG | LabelMe-VGG | MNIST-gray | Places205-alexnet
------------ | ------------- | ------------- | ------------- | ------------- | ------------ | ----------- | -------------
16 bit | 17.3 | 53.1 | 15.6 | 29.1 | 20.1 | 47.3 | 7.7
24 bit | 18.1 | 59.6 | 16.0 | 30.3 | 22.0 | 52.7 | 10.1
32 bit | 18.8 | 61.1 | 16.4 | 30.8 | 23.8 | 53.3 | 12.2
64 bit | 19.5 | 65.6 | 17.5 | 32.6 | 26.2 | 56.0 | 15.9
128 bit | 20.0 | 68.3 | 18.1 | 33.7 | 28.2 | 58.3 | 17.9

For more detailed results and experimental settings, please refer to our paper.

### Dataset Links:
OneDrive | gist | vggfc7 | uint | alexnet
------------ | ------------- | ------------- | ------------- | -------------
Cifar10 | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gnrlULhhHGy4Q88c)  | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gnwSyLlwvS29H7jX)  | ☐  | ☐ 
MNIST |[✓](https://1drv.ms/u/s!Av1MQK8mV3J8gnkooTeL9ZdtCYtu) | ☐  | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gn1NqWh31pMiZ0xs) | ☐ 
Labelme | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gnu2gPoTrm86tMqH) | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gn5dcG7GrQuktVoC) | ☐  | ☐ 
Places205 | ☐  | ☐  | ☐  | [✓](https://1drv.ms/u/s!Av1MQK8mV3J8gwCewchoZsCs3RZl)

Baidu Pan | gist | vggfc7 | uint | alexnet
------------ | ------------- | ------------- | ------------- | -------------
Cifar10 | [✓](https://pan.baidu.com/s/1nh-1F3imSeG7KErYWp5ERQ) (irou)  | [✓](https://pan.baidu.com/s/14YyiDcgSX_0zsewwpuVkFg) (0hr5) | ☐  | ☐ 
MNIST |[✓](https://pan.baidu.com/s/1C6dct3FvRLw1W7eM3kcSGQ) (6d29) | ☐  | [✓](https://pan.baidu.com/s/1H0qRPFWvlUfU7KGmOM1GYw) (masv) | ☐ 
Labelme | [✓](https://pan.baidu.com/s/1jddzOdGIzPcSzxIxsC2L2A) (47ny) | [✓](https://pan.baidu.com/s/11gFaGeu0sljMgjCdGWn4Og) (b7na) | ☐  | ☐ 
Places205 | ☐  | ☐  | ☐  | [✓](https://pan.baidu.com/s/1rvKXkl6amuwhjsf2I_s8NQ) (opw9)

`uint` refers to MNIST 784-D (28x28) gray-scale feature vector, which is saved as uint8.

### Brief Intro :
<img src="./img/KNNH.png" width="700" height="240" />

MSE may not lead to the optimal binary representation. We propose to use the conditional entropy as the criterion to refine the previous method.

### Q&A

1. By adjusting the hyper-parameters such as `KNNHparam.K` and `KNNHparam.times`, I get a better or worse result.

It is possible that our setting is suboptimal, which has been discussed in our paper (see Figure 3). We **did not** try to fine-tune those settings to overcome previous well-known methods.

2. Why KNNH performs poorly on `precision(%)@r=2`?

Precision@r is different from the well-known R-precision which describes one point on the precision-recall curve (we use p@r=2 because it is popular in recent deep hashings). In our experiments, though the precision@r=2 values by KNNH are not very good, the recall@r=2 of KNNH is higher than baselines. As we all know, it is common to compare the precision of two points on different PR curves with the same recall. Otherwise, the comparison is unfair.

### Acknowlegment :
We would like to thank the authors of [Gemb](https://github.com/hnanhtuan/Gemb) and [MiHash](https://github.com/fcakir/mihash) for sharing their codes! This project is built on previous well-known methods such as [ITQ](http://www.cs.unc.edu/~lazebnik/publications/cvpr11_small_code.pdf), [BA](https://arxiv.org/abs/1501.00756), [KMH](http://kaiminghe.com/publications/cvpr13kmh.pdf), [SH](https://papers.nips.cc/paper/3383-spectral-hashing), [PCAH](http://www.ee.columbia.edu/ln/dvmm/publications/12/PAMI_SSHASH.pdf) and [SPH](https://sglab.kaist.ac.kr/Spherical_Hashing/Spherical_Hashing.pdf).

### References :
```bib
@InProceedings{KNNH,
  author    = {Xiangyu He and Peisong Wang and Jian Cheng},
  title     = {K-Nearest Neighbors Hashing},
  booktitle = {2019 {IEEE} Conference on Computer Vision and Pattern Recognition},
  month     = {July},
  year      = {2019}
}
```

K-Nearest-Neighbors-Hashing

K-Nearest-Neighbors-Hashing 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

K-Nearest-Neighbors-Hashing

K-Nearest-Neighbors-Hashing 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README