A python implementation of the paper "Scalable Recognition with a Vocabulary Tree, D. Nister, H. Stewenius, 2006"

Home Page: http://www-inst.eecs.berkeley.edu/~cs294-6/fa06/papers/nister_stewenius_cvpr2006.pdf

Jupyter Notebook 83.21% Python 16.79%

deep-learning image-retrieval information-retrieval vocabulary-tree alexnet sift orb pytorch content-based-image-retrieval python

scalable-recognition-with-a-vocabulary-tree's Introduction

Scalable Recognition with a Vocabulary Tree

The code provided in this repository has been developed for teaching purposes at the Imperial College London. It is part of the Computer Vision Day of the Business School Executive Education Program for Sberbank.

Contributors

Getting started

1. Install the conda environment

A. If you are on windows

conda env create -f env\sberbank_win.yml

B. If you are on macOS or on linux platforms

conda env create -f env\sberbank_unix.yml

2. Start jupyter and open the notebook

conda activate cbir
jupyter lab

3. Open a terminal from jupyter and type

python cbir/download.py

Acknowledgements

The authors acknowledge the Executive Education of the Business School at the Imperial College for the support. We thank Professor Anil Bharath of the Department of Bioengineering for the guidance and the opportunity of being part of the Computer Vision Day. Thanks to Kai Arulkumaran and to Stathi Fotiadis for the feedback before the session and the assistance in teaching the session (2020).

Literature

Datasets:

Database indexing:

Features extraction:

End-to-end

Surveys

Code

scalable-recognition-with-a-vocabulary-tree's People

Contributors

Stargazers

Watchers

Forkers

lanckie liuchu1997 pengshengbin inkyusa imperial-college-for-sberbank liukelinlin yongduek jangsoopark

scalable-recognition-with-a-vocabulary-tree's Issues

Setup multiple ways to deploy the notebook

We will setup all the following modalities and use them in order of priority:

Ask to clone this repository
Send a compressed archive of this repository via email
Setup a google colab

Add extrac stage for data after download

Implement SIFT feature extraction

Consider using torch-sift, cv2 or numpy-sift.

This includes setting the repository and how we include the code from these libraries

Write a function that reads the siftgeo

Implement vocabulary tree

Add session timeline

I would say we split the session in parts and assign a time to each part.
Given a total of 90 minutes, we plan for 80 minutes and leave a 10 minutes buffer:

Time 00:10: 10 minutes setup, problems, contingencies
Time 00:15: 5 minutes describing the problem
Time 00:40: 25 minutes SIFT
Time 00:65: 25 minutes Vocabulary tree
Time 00:80: 15 minutes DCNN for features extraction

@RPFeynman what do you think?

Fine tuning of MSER

I'm struggling to find a set of parameters for MSER which makes it work well on natural scenes and "artificial" ones (like houses, indoor, text, etc). I wonder how much fine-tuning such method needs.

In any case, once the entire pipeline is ready we should do this.

Set python version in yml

A reminder to fix the yml file with the correct python version.

Pytorch can't be installed on python 3.10, which is the default now if the yml file is used to construct the environment (see pytorch/pytorch#66424 )

Error:

$ conda env create -f env/sberbank_unix.yml 
Collecting package metadata (repodata.json): done
Solving environment: done


==> WARNING: A newer version of conda exists. <==
  current version: 4.10.1
  latest version: 4.11.0

Please update conda by running

    $ conda update -n base -c defaults conda



Downloading and Extracting Packages
fontconfig-2.12.6    | 221 KB    | ##################################### | 100% 
pango-1.42.0         | 458 KB    | ##################################### | 100% 
pygraphviz-1.7       | 181 KB    | ##################################### | 100% 
pcre-8.45            | 207 KB    | ##################################### | 100% 
freetype-2.8         | 542 KB    | ##################################### | 100% 
jpeg-9d              | 232 KB    | ##################################### | 100% 
expat-2.4.1          | 168 KB    | ##################################### | 100% 
harfbuzz-1.7.6       | 474 KB    | ##################################### | 100% 
glib-2.69.1          | 1.7 MB    | ##################################### | 100% 
zstd-1.4.9           | 480 KB    | ##################################### | 100% 
pixman-0.40.0        | 373 KB    | ##################################### | 100% 
graphviz-2.40.1      | 6.5 MB    | ##################################### | 100% 
lz4-c-1.9.3          | 185 KB    | ##################################### | 100% 
cairo-1.14.12        | 905 KB    | ##################################### | 100% 
libxml2-2.9.12       | 1.2 MB    | ##################################### | 100% 
certifi-2021.5.30    | 148 KB    | ##################################### | 100% 
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
Installing pip dependencies: \ Ran pip subprocess with arguments:
['/home/antonio/anaconda3/envs/cbir/bin/python', '-m', 'pip', 'install', '-U', '-r', '/home/antonio/repos/scalable-recognition-with-a-vocabulary-tree/env/condaenv.u72yw4q5.requirements.txt']
Pip subprocess output:
Collecting future
  Downloading future-0.18.2.tar.gz (829 kB)
Collecting h5py
  Downloading h5py-3.6.0-cp310-cp310-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (4.5 MB)
Collecting jupyterlab
  Using cached jupyterlab-3.2.8-py3-none-any.whl (8.5 MB)
Collecting ipywidgets
  Using cached ipywidgets-7.6.5-py2.py3-none-any.whl (121 kB)
Collecting matplotlib
  Using cached matplotlib-3.5.1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (11.9 MB)
Collecting termcolor
  Using cached termcolor-1.1.0.tar.gz (3.9 kB)
Collecting networkx
  Using cached networkx-2.6.3-py3-none-any.whl (1.9 MB)
Collecting pandas
  Using cached pandas-1.3.5-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (11.5 MB)

Pip subprocess error:
ERROR: Could not find a version that satisfies the requirement torch (from versions: none)
ERROR: No matching distribution found for torch

failed

CondaEnvException: Pip failed

Define system requirements for the notebook

Outcome: a document with all the required configurations, and an email to Inka 2 weeks before the start of the session.

Make sure the download function has a total_size

At the moment the ftp server does not provide the size of the files to download.

We need to hard code the sizes

epignatelli / scalable-recognition-with-a-vocabulary-tree Goto Github PK

scalable-recognition-with-a-vocabulary-tree's Introduction

Scalable Recognition with a Vocabulary Tree

Contributors

Getting started

1. Install the conda environment

2. Start jupyter and open the notebook

3. Open a terminal from jupyter and type

Acknowledgements

Literature

Datasets:

Database indexing:

Features extraction:

End-to-end

Surveys

Code

scalable-recognition-with-a-vocabulary-tree's People

Contributors

Stargazers

Watchers

Forkers

scalable-recognition-with-a-vocabulary-tree's Issues

Recommend Projects

Recommend Topics

Recommend Org