DeepScores-A Dataset for Segmentation, Detection and Classification of Tiny Objects.
Lukas TuggenerIsmail EleziJürgen SchmidhuberMarcello PelilloThilo StadelmannPublished in: ICPR (2018)
Keyphrases
- feature extraction and classification
- bounding box
- features extraction
- foreground background separation
- object segmentation
- classification accuracy
- complex background
- multiple objects
- benchmark datasets
- image segmentation
- detecting objects
- cluttered scenes
- decision trees
- feature extraction
- support vector
- page segmentation
- image classification
- segmentation algorithm
- medical images
- pixel classification
- cluttered background
- detection algorithm
- false positives
- partial occlusion
- object detection
- category labels
- lung nodules
- discriminative classifiers
- video scene
- image segments
- edge detection
- sign recognition
- text classification
- outdoor images
- pigmented skin lesions
- feature selection
- feature space
- feature vectors
- man made
- outdoor scenes
- support vector machine
- image features
- shape prior
- segmentation method
- feature set
- face detection