This is a set of scripts for creating visual summary for all papers published at CVPR. The original idea is from https://github.com/colincsl/cvpr2013papers They show how one can manipulate PDFs, analyze word frequencies,
-
Clone this repository
git clone https://github.com/xiangze/cvprpapers.git -
Make sure you have the following libraries [System: ghostscript, imagemagick; Python: httplib2, beautifulsoup]
pip3 install -r requirements.txt
On OSX: brew install ghostscript brew install imagemagick pip install httplib2 pip install beautifulsoup pip install gensim pip install pyLDAvis pip install nltk
1.1 Download NLTK stopwords
>>import nltk
>>nltk.download("stopwords")for year in `seq 2016 2023`;do
./run.sh cvpr $year
done
contents of run.sh
| python script | from | to |
|---|---|---|
| download_paper.py | cvpr+"year" | $conference$year/*.pdf |
| pdftowordcloud.py | $conference$year/*.pdf | topwords.p |
| scrape_pdfs.py | $conference$year/*.pdf | all_text.txt |
| makecorpus.py | $conference$year/*.pdf | alpapers.txt |
| genLDA.py | allpapers.txt | papers_%d.model' |
| genpages.py | papers_model:lda | $conference$year_$topicnum.html |
WTFPL licence