A5055189380- Handwritten Text Recognition Techniques
- Image Processing and 3D Reconstruction
- Natural Language Processing Techniques
- Vehicle License Plate Recognition
- Digital and Cyber Forensics
- Multimodal Machine Learning Applications
- Advanced Image and Video Retrieval Techniques
- Topic Modeling
- Neural Networks and Applications
- Model Reduction and Neural Networks
- Text and Document Classification Technologies
- Railway Systems and Energy Efficiency
- Advanced Control Systems Optimization
- semigroups and automata theory
- Digital and Traditional Archives Management
- Electric Vehicles and Infrastructure
- Electric and Hybrid Vehicle Technologies
- Algorithms and Data Compression
- Maritime Transport Emissions and Efficiency
- Aerospace Engineering and Control Systems
- 3D Surveying and Cultural Heritage
- Computational Physics and Python Applications
- Image Retrieval and Classification Techniques
- Explainable Artificial Intelligence (XAI)
- Digital Humanities and Scholarship
Planet
2019-2022
University of Rostock
2012-2019
Encoder-decoder models have become an effective approach for sequence learning tasks like machine translation, image captioning and speech recognition, but yet to show competitive results handwritten text recognition. To this end, we propose attention-based sequence-to-sequence model. It combines a convolutional neural network as generic feature extractor with recurrent encode both the visual information, well temporal context between characters in input image, uses separate decode actual...
Purpose An overview of the current use handwritten text recognition (HTR) on archival manuscript material, as provided by EU H2020 funded Transkribus platform. It explains HTR, demonstrates , gives examples cases, highlights affect HTR may have scholarship, and evidences this turning point advanced digitised heritage content. The paper aims to discuss these issues. Design/methodology/approach This adopts a case study approach, using development delivery one openly available platform for...
The cBAD competition aims at benchmarking state-of-the-art baseline detection algorithms. It is in line with previous competitions such as the ICDAR 2013 Handwriting Segmentation Contest. A new, challenging, dataset was created to test behavior of systems on real world data. Since traditional evaluation schemes are not applicable size and modality this dataset, we present a new one that introduces baselines measure performance. We received submissions from five different teams for both tracks.
Text line detection is crucial for any application associated with Automatic Recognition or Keyword Spotting. Modern algorithms perform good on well-established datasets since they either comprise clean data simple/homogeneous page layouts. We have collected and annotated 2036 archival document images from different locations time periods. The dataset contains varying layouts degradations that challenge text segmentation methods. Well established evaluation schemes such as the Detection Rate...
The transcription of handwritten text on images is one task in machine learning and solution to solve it using multi-dimensional recurrent neural networks (MDRNN) with connectionist temporal classification (CTC). RNNs can contain special units, the long short-term memory (LSTM) cells. They are able learn term dependencies but they get unstable when dimension chosen greater than one. We defined some useful necessary properties for one-dimensional LSTM cell extend them case. Thereby we...
We describe CITlab's recognition system for the HTRtS competition attached to 13. International Conference on Document Analysis and Recognition, ICDAR 2015. The task comprises of historical handwritten documents. core algorithms our are based multi-dimensional recurrent neural networks (MDRNN) connectionist temporal classification (CTC). software modules behind that as well basic utility technologies essentially powered by PLANET's ARGUS framework intelligent text image processing.
In this paper, a basic simulation model of the longitudinal dynamics is developed for diesel hybrid railway vehicle with battery as energy storage system. This represents an inverse system description that employs given duty cycle to compute all variables in purely algebraic manner. Based on model, three different optimization strategies aiming at minimization fuel consumption are investigated. The parameters used these approaches allow modification certain phases operating strategy. results...
To calculate feedforward control strategies by means of dynamic optimization procedures, the alternatives direct and indirect methods are compared in this paper. The approaches that considered paper make use Hermite-Simpson method Legendre Pseudospectral as underlying discretization strategies. Their description is followed a short introduction to maximum principle Pontryagin, i.e., method. All procedures employed compute sequences for flight four-rotor UAV, which an unstable, nonlinear...
This paper presents numerical results for the optimization of trajectories standard diesel and hybrid railway vehicles with respect to fuel consumption between two successive stops. First, a nonlinear model complete power train vehicle, purely driven by an internal combustion engine, is described. Based on four different operating regimes - acceleration, cruising, coasting, braking fuel-optimal are computed determining optimal regime sequence. For this purpose, golden ratio search method...
tfaip is a Python-based research framework for developing, structuring, and deploying Deep Learning projects powered by Tensorflow (Abadi et al., 2015) intended scientists of universities or organizations who research, develop, optionally deploy models.tfaip enables both simple complex implementation scenarios, such as image classification, object detection, text recognition, natural language processing, speech recognition.Each scenario highly configurable parameters that can directly be...
We describe CITlab's recognition system for the ANWRESH-2014 competition attached to 14. International Conference on Frontiers in Handwriting Recognition, ICFHR 2014. The task comprises word from segmented historical documents. core components of our are based multi-dimensional recurrent neural networks (MDRNN) and connectionist temporal classification (CTC). software modules behind that as well basic utility technologies essentially powered by PLANET's ARGUS framework intelligent text image...
Measuring the performance of text recognition and line detection engines is an important step to objectively compare systems their configuration. There exist well-established measures for both tasks separately. However, there no sophisticated evaluation scheme measure quality a combined system. The F-measure on word level well-known methodology, which sometimes used in this context. Nevertheless, it does not take into account alignment hypothesis ground truth can lead deceptive results....