jimnoneill commited on
Commit
9b4dd95
·
verified ·
1 Parent(s): b5b458d

Upload src/poster_sentry/__init__.py with huggingface_hub

Browse files
Files changed (1) hide show
  1. src/poster_sentry/__init__.py +24 -0
src/poster_sentry/__init__.py ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ """
2
+ PosterSentry — Multimodal Scientific Poster Classifier
3
+ =======================================================
4
+
5
+ Classifies PDFs as scientific posters vs. non-posters using both
6
+ text embeddings (model2vec) and visual features (image analysis).
7
+
8
+ Trained on 30K+ real scientific posters from Zenodo and Figshare
9
+ via the posters.science initiative at FAIR Data Innovations Hub.
10
+
11
+ Usage:
12
+ from poster_sentry import PosterSentry
13
+
14
+ sentry = PosterSentry()
15
+ sentry.initialize()
16
+ result = sentry.classify("document.pdf")
17
+ # {'is_poster': True, 'confidence': 0.97, 'text_score': 0.95, 'visual_score': 0.99}
18
+ """
19
+
20
+ from .classifier import PosterSentry
21
+ from .features import VisualFeatureExtractor, PDFStructuralExtractor
22
+
23
+ __version__ = "0.1.0"
24
+ __all__ = ["PosterSentry", "VisualFeatureExtractor", "PDFStructuralExtractor"]