activity
20172021
most citedHistorical Document Image Segmentation with LDA-Initialized Deep Neural Networks

25 citations · 25 across the 4 of their papers we have counts for

collaborators

11 papers

cs.CV2021

Generating Synthetic Handwritten Historical Documents With OCR Constrained GANs

Lars Vögtlin, Manuel Drazyk, Vinaychandran Pondenkandath +2

We present a framework to generate synthetic historical documents with precise ground truth using nothing more than a collection of unlabeled historical images. Obtaining large lab…

cs.CV2019

Labeling, Cutting, Grouping: an Efficient Text Line Segmentation Method for Medieval Manuscripts

Michele Alberti, Lars Vögtlin, Vinaychandran Pondenkandath +3

This paper introduces a new way for text-line extraction by integrating deep-learning based pre-classification and state-of-the-art segmentation methods. Text-line extraction in co…

cs.LG2019

Improving Reproducible Deep Learning Workflows with DeepDIVA

Michele Alberti, Vinaychandran Pondenkandath, Lars Vögtlin +3

The field of deep learning is experiencing a trend towards producing reproducible research. Nevertheless, it is still often a frustrating experience to reproduce scientific results…

cs.AI2019

Survey of Artificial Intelligence for Card Games and Its Application to the Swiss Game Jass

Joel Niklaus, Michele Alberti, Vinaychandran Pondenkandath +2

In the last decades we have witnessed the success of applications of Artificial Intelligence to playing games. In this work we address the challenging field of games with hidden in…

cs.CV2019

A Comprehensive Study of ImageNet Pre-Training for Historical Document Image Analysis

Linda Studer, Michele Alberti, Vinaychandran Pondenkandath +5

Automatic analysis of scanned historical documents comprises a wide range of image analysis tasks, which are often challenging for machine learning due to a lack of human-annotated…

cs.LG2018

Leveraging Random Label Memorization for Unsupervised Pre-Training

Vinaychandran Pondenkandath, Michele Alberti, Sammer Puran +2

We present a novel approach to leverage large unlabeled datasets by pre-training state-of-the-art deep neural networks on randomly-labeled datasets. Specifically, we train the neur…