Vincent Christlein

Konstantina Nikolaidou, George Retsinas, Vincent Christlein et al.

Text-to-Image synthesis is the task of generating an image according to a specific text description. Generative Adversarial Networks have been considered the standard method for image synthesis virtually since their introduction. Denoising Diffusion Probabilistic Models are recently setting a new baseline, with remarkable results in Text-to-Image synthesis, among other fields. Aside its usefulness per se, it can also be particularly relevant as a tool for data augmentation to aid training models for other document image processing tasks. In this work, we present a latent diffusion-based method for styled text-to-text-content-image generation on word-level. Our proposed method is able to generate realistic word image samples from different writer styles, by using class index styles and text content prompts without the need of adversarial training, writer recognition, or text recognition. We gauge system performance with the Fréchet Inception Distance, writer recognition accuracy, and writer retrieval. We show that the proposed model produces samples that are aesthetically pleasing, help boosting text recognition performance, and get similar writer retrieval score as real data. Code is available at: https://github.com/koninik/WordStylist.

Fei Wu, Nora Gourmelon, Thorsten Seehaus et al.

Knowledge on changes in glacier calving front positions is important for assessing the status of glaciers. Remote sensing imagery provides the ideal database for monitoring calving front positions, however, it is not feasible to perform this task manually for all calving glaciers globally due to time-constraints. Deep learning-based methods have shown great potential for glacier calving front delineation from optical and radar satellite imagery. The calving front is represented as a single thin line between the ocean and the glacier, which makes the task vulnerable to inaccurate predictions. The limited availability of annotated glacier imagery leads to a lack of data diversity (not all possible combinations of different weather conditions, terminus shapes, sensors, etc. are present in the data), which exacerbates the difficulty of accurate segmentation. In this paper, we propose Attention-Multi-hooking-Deep-supervision HookNet (AMD-HookNet), a novel glacier calving front segmentation framework for synthetic aperture radar (SAR) images. The proposed method aims to enhance the feature representation capability through multiple information interactions between low-resolution and high-resolution inputs based on a two-branch U-Net. The attention mechanism, integrated into the two branch U-Net, aims to interact between the corresponding coarse and fine-grained feature maps. This allows the network to automatically adjust feature relationships, resulting in accurate pixel-classification predictions. Extensive experiments and comparisons on the challenging glacier segmentation benchmark dataset CaFFe show that our AMD-HookNet achieves a mean distance error of 438 m to the ground truth outperforming the current state of the art by 42%, which validates its effectiveness.

Mathias Zinnen, Prathmesh Madhu, Ronak Kosti et al.

The Odeuropa Challenge on Olfactory Object Recognition aims to foster the development of object detection in the visual arts and to promote an olfactory perspective on digital heritage. Object detection in historical artworks is particularly challenging due to varying styles and artistic periods. Moreover, the task is complicated due to the particularity and historical variance of predefined target objects, which exhibit a large intra-class variance, and the long tail distribution of the dataset labels, with some objects having only very few training examples. These challenges should encourage participants to create innovative approaches using domain adaptation or few-shot learning. We provide a dataset of 2647 artworks annotated with 20 120 tightly fit bounding boxes that are split into a training and validation set (public). A test set containing 1140 artworks and 15 480 annotations is kept private for the challenge evaluation.

Anguelos Nicolaou, Vincent Christlein, Edgar Riba et al.

We propose the use of fractals as a means of efficient data augmentation. Specifically, we employ plasma fractals for adapting global image augmentation transformations into continuous local transforms. We formulate the diamond square algorithm as a cascade of simple convolution operations allowing efficient computation of plasma fractals on the GPU. We present the TorMentor image augmentation framework that is totally modular and deterministic across images and point-clouds. All image augmentation operations can be combined through pipelining and random branching to form flow networks of arbitrary width and depth. We demonstrate the efficiency of the proposed approach with experiments on document image segmentation (binarization) with the DIBCO datasets. The proposed approach demonstrates superior performance to traditional image augmentation techniques. Finally, we use extended synthetic binary text images in a self-supervision regiment and outperform the same model when trained with limited data and simple extensions.

Aline Sindel, Bettina Hohberger, Andreas Maier et al.

In ophthalmological imaging, multiple imaging systems, such as color fundus, infrared, fluorescein angiography, optical coherence tomography (OCT) or OCT angiography, are often involved to make a diagnosis of retinal disease. Multi-modal retinal registration techniques can assist ophthalmologists by providing a pixel-based comparison of aligned vessel structures in images from different modalities or acquisition times. To this end, we propose an end-to-end trainable deep learning method for multi-modal retinal image registration. Our method extracts convolutional features from the vessel structure for keypoint detection and description and uses a graph neural network for feature matching. The keypoint detection and description network and graph neural network are jointly trained in a self-supervised manner using synthetic multi-modal image pairs and are guided by synthetically sampled ground truth homographies. Our method demonstrates higher registration accuracy as competing methods for our synthetic retinal dataset and generalizes well for our real macula dataset and a public fundus dataset.

Mathias Zinnen, Azhar Hussian, Hang Tran et al.

Smell gestures play a crucial role in the investigation of past smells in the visual arts yet their automated recognition poses significant challenges. This paper introduces the SniffyArt dataset, consisting of 1941 individuals represented in 441 historical artworks. Each person is annotated with a tightly fitting bounding box, 17 pose keypoints, and a gesture label. By integrating these annotations, the dataset enables the development of hybrid classification approaches for smell gesture recognition. The datasets high-quality human pose estimation keypoints are achieved through the merging of five separate sets of keypoint annotations per person. The paper also presents a baseline analysis, evaluating the performance of representative algorithms for detection, keypoint estimation, and classification tasks, showcasing the potential of combining keypoint estimation with smell gesture classification. The SniffyArt dataset lays a solid foundation for future research and the exploration of multi-task approaches leveraging pose keypoints and person boxes to advance human gesture and olfactory dimension analysis in historical artworks.

Vincent Christlein, Isabelle Marthot-Santaniello, Martin Mayr et al.

The analysis of digitized historical manuscripts is typically addressed by paleographic experts. Writer identification refers to the classification of known writers while writer retrieval seeks to find the writer by means of image similarity in a dataset of images. While automatic writer identification/retrieval methods already provide promising results for many historical document types, papyri data is very challenging due to the fiber structures and severe artifacts. Thus, an important step for an improved writer identification is the preprocessing and feature sampling process. We investigate several methods and show that a good binarization is key to an improved writer identification in papyri writings. We focus mainly on writer retrieval using unsupervised feature methods based on traditional or self-supervised-based methods. It is, however, also comparable to the state of the art supervised deep learning-based method in the case of writer classification/re-identification.

Aline Sindel, Andreas Maier, Vincent Christlein

Facial landmark detection plays an important role for the similarity analysis in artworks to compare portraits of the same or similar artists. With facial landmarks, portraits of different genres, such as paintings and prints, can be automatically aligned using control-point-based image registration. We propose a deep-learning-based method for facial landmark detection in high-resolution images of paintings and prints. It divides the task into a global network for coarse landmark prediction and multiple region networks for precise landmark refinement in regions of the eyes, nose, and mouth that are automatically determined based on the predicted global landmark coordinates. We created a synthetically augmented facial landmark art dataset including artistic style transfer and geometric landmark shifts. Our method demonstrates an accurate detection of the inner facial landmarks for our high-resolution dataset of artworks while being comparable for a public low-resolution artwork dataset in comparison to competing methods.

Mathias Zinnen, Prathmesh Madhu, Peter Bell et al.

We investigate the effect of style and category similarity in multiple datasets used for object detection pretraining. We find that including an additional stage of object-detection pretraining can increase the detection performance considerably. While our experiments suggest that style similarities between pre-training and target datasets are less important than matching categories, further experiments are needed to verify this hypothesis.

Martin Mayr, Marcel Dreier, Florian Kordon et al.

The imitation of cursive handwriting is mainly limited to generating handwritten words or lines. Multiple synthetic outputs must be stitched together to create paragraphs or whole pages, whereby consistency and layout information are lost. To close this gap, we propose a method for imitating handwriting at the paragraph level that also works for unseen writing styles. Therefore, we introduce a modified latent diffusion model that enriches the encoder-decoder mechanism with specialized loss functions that explicitly preserve the style and content. We enhance the attention mechanism of the diffusion model with adaptive 2D positional encoding and the conditioning mechanism to work with two modalities simultaneously: a style image and the target text. This significantly improves the realism of the generated handwriting. Our approach sets a new benchmark in our comprehensive evaluation. It outperforms all existing imitation methods at both line and paragraph levels, considering combined style and content preservation.

Aline Sindel, Andreas Maier, Vincent Christlein

For art investigations of paintings, multiple imaging technologies, such as visual light photography, infrared reflectography, ultraviolet fluorescence photography, and x-radiography are often used. For a pixel-wise comparison, the multi-modal images have to be registered. We present a registration and visualization software tool, that embeds a convolutional neural network to extract cross-modal features of the crack structures in historical paintings for automatic registration. The graphical user interface processes the user's input to configure the registration parameters and to interactively adapt the image views with the registered pair and image overlays, such as by individual or synchronized zoom or movements of the views. In the evaluation, we qualitatively and quantitatively show the effectiveness of our software tool in terms of registration performance and short inference time on multi-modal paintings and its transferability by applying our method to historical prints.

Tobias Koch, Felix Liebezeit, Christian Riess et al.

Dynamic environments require adaptive applications. One particular machine learning problem in dynamic environments is open world recognition. It characterizes a continuously changing domain where only some classes are seen in one batch of the training data and such batches can only be learned incrementally. Open world recognition is a demanding task that is, to the best of our knowledge, addressed by only a few methods. This work introduces a modification of the widely known Extreme Value Machine (EVM) to enable open world recognition. Our proposed method extends the EVM with a partial model fitting function by neglecting unaffected space during an update. This reduces the training time by a factor of 28. In addition, we provide a modified model reduction using weighted maximum K-set cover to strictly bound the model complexity and reduce the computational effort by a factor of 3.5 from 2.1 s to 0.6 s. In our experiments, we rigorously evaluate openness with two novel evaluation protocols. The proposed method achieves superior accuracy of about 12 % and computational efficiency in the tasks of image classification and face recognition.

Martin Röhn, Nora Gourmelon, Vincent Christlein

Climate adaptation is vital for the sustainability and sometimes the mere survival of our urban areas. However, small cities often struggle with limited personnel resources and integrating vast amounts of data from multiple sources for a comprehensive analysis. To overcome these challenges, this paper proposes a multi-layered system combining specialized LLMs, satellite imagery analysis and a knowledge base to aid in developing effective climate adaptation strategies. The corresponding code can be found at https://github.com/Photon-GitHub/EcoScapes.

Aline Sindel, Andreas Maier, Vincent Christlein

Unpaired image-to-image translation of retinal images can efficiently increase the training dataset for deep-learning-based multi-modal retinal registration methods. Our method integrates a vessel segmentation network into the image-to-image translation task by extending the CycleGAN framework. The segmentation network is inserted prior to a UNet vision transformer generator network and serves as a shared representation between both domains. We reformulate the original identity loss to learn the direct mapping between the vessel segmentation and the real image. Additionally, we add a segmentation loss term to ensure shared vessel locations between fake and real images. In the experiments, our method shows a visually realistic look and preserves the vessel structures, which is a prerequisite for generating multi-modal training data for image registration.

Anguelos Nicolaou, Daniel Luger, Franziska Decker et al.

Diplomatics, the analysis of medieval charters, is a major field of research in which paleography is applied. Annotating data, if performed by laymen, needs validation and correction by experts. In this paper, we propose an effective and efficient annotation approach for charter segmentation, essentially reducing it to object detection. This approach allows for a much more efficient use of the paleographer's time and produces results that can compete and even outperform pixel-level segmentation in some use cases. Further experiments shed light on how to design a class ontology in order to make the best use of annotators' time and effort. Exploiting the presence of calibration cards in the image, we further annotate the data with the physical length in pixels and train regression neural networks to predict it from image patches.

Prathmesh Madhu, Tilman Marquart, Ronak Kosti et al.

Image compositions are helpful in the study of image structures and assist in discovering the semantics of the underlying scene portrayed across art forms and styles. With the digitization of artworks in recent years, thousands of images of a particular scene or narrative could potentially be linked together. However, manually linking this data with consistent objectiveness can be a highly challenging and time-consuming task. In this work, we present a novel approach called Image Composition Canvas (ICC++) to compare and retrieve images having similar compositional elements. ICC++ is an improvement over ICC specializing in generating low and high-level features (compositional elements) motivated by Max Imdahl's work. To this end, we present a rigorous quantitative and qualitative comparison of our approach with traditional and state-of-the-art (SOTA) methods showing that our proposed method outperforms all of them. In combination with deep features, our method outperforms the best deep learning-based method, opening the research direction for explainable machine learning for digital humanities. We will release the code and the data post-publication.

Vishal Patoliya, Mathias Zinnen, Andreas Maier et al.

Emotions and smell are underrepresented in digital art history. In this exploratory work, we show that recognising emotions from smell-related artworks is technically feasible but has room for improvement. Using style transfer and hyperparameter optimization we achieve a minor performance boost and open up the field for future extensions.

Marcel Dreier, Nora Gourmelon, Dakota Pyles et al.

During benchmarking, the state-of-the-art model for glacier calving front delineation achieves near-human performance. However, when applied in a real-world setting at a novel study site, its delineation accuracy is insufficient for calving front products intended for further scientific analyses. This site represents an out-of-distribution domain for a model trained solely on the benchmark dataset. By employing a few-shot domain adaptation strategy, incorporating spatial static prior knowledge, and including summer reference images in the input time series, the delineation error is reduced from 1131.6 m to 68.7 m without any architectural modifications. These methodological advancements establish a framework for applying deep learning-based calving front segmentation to novel study sites, enabling calving front monitoring on a global scale.

Shumei Liu, Haiting Huang, Mathias Zinnen et al.

Olfaction, often overlooked in cultural heritage studies, holds profound significance in shaping human experiences and identities. Examining historical depictions of olfactory scenes can offer valuable insights into the role of smells in history. We show that a transfer-learning approach using weakly labeled training data can remarkably improve the classification of fragrant spaces and, more generally, artistic scene depictions. We fine-tune Places365-pre-trained models by querying two cultural heritage data sources and using the search terms as supervision signal. The models are evaluated on two manually corrected test splits. This work lays a foundation for further exploration of fragrant spaces recognition and artistic scene classification. All images and labels are released as the ArtPlaces dataset at https://zenodo.org/doi/10.5281/zenodo.11584328.

Richin Sukesh, Mathias Seuret, Anguelos Nicolaou et al.

Binarization of document images is an important pre-processing step in the field of document analysis. Traditional image binarization techniques usually rely on histograms or local statistics to identify a valid threshold to differentiate between different aspects of the image. Deep learning techniques are able to generate binarized versions of the images by learning context-dependent features that are less error-prone to degradation typically occurring in document images. In recent years, many deep learning-based methods have been developed for document binarization. But which one to choose? There have been no studies that compare these methods rigorously. Therefore, this work focuses on the evaluation of different deep learning-based methods under the same evaluation protocol. We evaluate them on different Document Image Binarization Contest (DIBCO) datasets and obtain very heterogeneous results. We show that the DE-GAN model was able to perform better compared to other models when evaluated on the DIBCO2013 dataset while DP-LinkNet performed best on the DIBCO2017 dataset. The 2-StageGAN performed best on the DIBCO2018 dataset while SauvolaNet outperformed the others on the DIBCO2019 challenge. Finally, we make the code, all models and evaluation publicly available (https://github.com/RichSu95/Document_Binarization_Collection) to ensure reproducibility and simplify future binarization evaluations.

Prathmesh Madhu, Tilman Marquart, Ronak Kosti et al.

Image compositions as a tool for analysis of artworks is of extreme significance for art historians. These compositions are useful in analyzing the interactions in an image to study artists and their artworks. Max Imdahl in his work called Ikonik, along with other prominent art historians of the 20th century, underlined the aesthetic and semantic importance of the structural composition of an image. Understanding underlying compositional structures within images is challenging and a time consuming task. Generating these structures automatically using computer vision techniques (1) can help art historians towards their sophisticated analysis by saving lot of time; providing an overview and access to huge image repositories and (2) also provide an important step towards an understanding of man made imagery by machines. In this work, we attempt to automate this process using the existing state of the art machine learning techniques, without involving any form of training. Our approach, inspired by Max Imdahl's pioneering work, focuses on two central themes of image composition: (a) detection of action regions and action lines of the artwork; and (b) pose-based segmentation of foreground and background. Currently, our approach works for artworks comprising of protagonists (persons) in an image. In order to validate our approach qualitatively and quantitatively, we conduct a user study involving experts and non-experts. The outcome of the study highly correlates with our approach and also demonstrates its domain-agnostic capability. We have open-sourced the code at https://github.com/image-compostion-canvas-group/image-compostion-canvas.

Prathmesh Madhu, Ronak Kosti, Lara Mührenberg et al.

In the field of Art History, images of artworks and their contexts are core to understanding the underlying semantic information. However, the highly complex and sophisticated representation of these artworks makes it difficult, even for the experts, to analyze the scene. From the computer vision perspective, the task of analyzing such artworks can be divided into sub-problems by taking a bottom-up approach. In this paper, we focus on the problem of recognizing the characters in Art History. From the iconography of $Annunciation$ $of$ $the$ $Lord$ (Figure 1), we consider the representation of the main protagonists, $Mary$ and $Gabriel$, across different artworks and styles. We investigate and present the findings of training a character classifier on features extracted from their face images. The limitations of this method, and the inherent ambiguity in the representation of $Gabriel$, motivated us to consider their bodies (a bigger context) to analyze in order to recognize the characters. Convolutional Neural Networks (CNN) trained on the bodies of $Mary$ and $Gabriel$ are able to learn person related features and ultimately improve the performance of character recognition. We introduce a new technique that generates more data with similar styles, effectively creating data in the similar domain. We present experiments and analysis on three different models and show that the model trained on domain related data gives the best performance for recognizing character. Additionally, we analyze the localized image regions for the network predictions. Code is open-sourced and available at https://github.com/prathmeshrmadhu/recognize_characters_art_history and the link to the published peer-reviewed article is https://dl.acm.org/citation.cfm?id=3357242.

Vincent Christlein, David Bernecker, Andreas Maier et al.

Convolutional neural networks (CNNs) have recently become the state-of-the-art tool for large-scale image classification. In this work we propose the use of activation features from CNNs as local descriptors for writer identification. A global descriptor is then formed by means of GMM supervector encoding, which is further improved by normalization with the KL-Kernel. We evaluate our method on two publicly available datasets: the ICDAR 2013 benchmark database and the CVL dataset. While we perform comparably to the state of the art on CVL, our proposed method yields about 0.21 absolute improvement in terms of mAP on the challenging bilingual ICDAR dataset.

Mathias Zinnen, Prathmesh Madhu, Inger Leemans et al.

Real-world applications of computer vision in the humanities require algorithms to be robust against artistic abstraction, peripheral objects, and subtle differences between fine-grained target classes. Existing datasets provide instance-level annotations on artworks but are generally biased towards the image centre and limited with regard to detailed object classes. The proposed ODOR dataset fills this gap, offering 38,116 object-level annotations across 4712 images, spanning an extensive set of 139 fine-grained categories. Conducting a statistical analysis, we showcase challenging dataset properties, such as a detailed set of categories, dense and overlapping objects, and spatial distribution over the whole image canvas. Furthermore, we provide an extensive baseline analysis for object detection models and highlight the challenging properties of the dataset through a set of secondary studies. Inspiring further research on artwork object detection and broader visual cultural heritage studies, the dataset challenges researchers to explore the intersection of object recognition and smell perception.

Alexander Schmidt, Prathmesh Madhu, Andreas Maier et al.

Image enhancement algorithms are very useful for real world computer vision tasks where image resolution is often physically limited by the sensor size. While state-of-the-art deep neural networks show impressive results for image enhancement, they often struggle to enhance real-world images. In this work, we tackle a real-world setting: inpainting of images from Dunhuang caves. The Dunhuang dataset consists of murals, half of which suffer from corrosion and aging. These murals feature a range of rich content, such as Buddha statues, bodhisattvas, sponsors, architecture, dance, music, and decorative patterns designed by different artists spanning ten centuries, which makes manual restoration challenging. We modify two different existing methods (CAR, HINet) that are based upon state-of-the-art (SOTA) super resolution and deblurring networks. We show that those can successfully inpaint and enhance these deteriorated cave paintings. We further show that a novel combination of CAR and HINet, resulting in our proposed inpainting network (ARIN), is very robust to external noise, especially Gaussian noise. To this end, we present a quantitative and qualitative comparison of our proposed approach with existing SOTA networks and winners of the Dunhuang challenge. One of the proposed methods HINet) represents the new state of the art and outperforms the 1st place of the Dunhuang Challenge, while our combination ARIN, which is robust to noise, is comparable to the 1st place. We also present and discuss qualitative results showing the impact of our method for inpainting on Dunhuang cave images.

Martin Mayr, Julian Krenz, Katharina Neumeier et al.

Most datasets in the field of document analysis utilize highly standardized labels, which, while simplifying specific tasks, often produce outputs that are not directly applicable to humanities research. In contrast, the Nuremberg Letterbooks dataset, which comprises historical documents from the early 15th century, addresses this gap by providing multiple types of transcriptions and accompanying metadata. This approach allows for developing methods that are more closely aligned with the needs of the humanities. The dataset includes 4 books containing 1711 labeled pages written by 10 scribes. Three types of transcriptions are provided for handwritten text recognition: Basic, diplomatic, and regularized. For the latter two, versions with and without expanded abbreviations are also available. A combination of letter ID and writer ID supports writer identification due to changing writers within pages. In the technical validation, we established baselines for various tasks, demonstrating data consistency and providing benchmarks for future research to build upon.

Nora Gourmelon, Konrad Heidler, Erik Loebel et al.

Calving front position variation of marine-terminating glaciers is an indicator of ice mass loss and a crucial parameter in numerical glacier models. Deep Learning (DL) systems can automatically extract this position from Synthetic Aperture Radar (SAR) imagery, enabling continuous, weather- and illumination-independent, large-scale monitoring. This study presents the first comparison of DL systems on a common calving front benchmark dataset. A multi-annotator study with ten annotators is performed to contrast the best-performing DL system against human performance. The best DL model's outputs deviate 221 m on average, while the average deviation of the human annotators is 38 m. This significant difference shows that current DL systems do not yet match human performance and that further research is needed to enable fully automated monitoring of glacier calving fronts. The study of Vision Transformers, foundation models, and the inclusion and processing strategy of more information are identified as avenues for future research.

Ahmed Sheta, Mathias Zinnen, Aline Sindel et al.

Finding smell references in historic artworks is a challenging problem. Beyond artwork-specific challenges such as stylistic variations, their recognition demands exceptionally detailed annotation classes, resulting in annotation sparsity and extreme class imbalance. In this work, we explore the potential of synthetic data generation to alleviate these issues and enable accurate detection of smell-related objects. We evaluate several diffusion-based augmentation strategies and demonstrate that incorporating synthetic data into model training can improve detection performance. Our findings suggest that leveraging the large-scale pretraining of diffusion models offers a promising approach for improving detection accuracy, particularly in niche applications where annotations are scarce and costly to obtain. Furthermore, the proposed approach proves to be effective even with relatively small amounts of data, and scaling it up provides high potential for further enhancements.

Nora Gourmelon, Marcel Dreier, Martin Mayr et al.

Glaciers are losing ice mass at unprecedented rates, increasing the need for accurate, year-round monitoring to understand frontal ablation, particularly the factors driving the calving process. Deep learning models can extract calving front positions from Synthetic Aperture Radar imagery to track seasonal ice losses at the calving fronts of marine- and lake-terminating glaciers. The current state-of-the-art model relies on ImageNet-pretrained weights. However, they are suboptimal due to the domain shift between the natural images in ImageNet and the specialized characteristics of remote sensing imagery, in particular for Synthetic Aperture Radar imagery. To address this challenge, we propose two novel self-supervised multimodal pretraining techniques that leverage SSL4SAR, a new unlabeled dataset comprising 9,563 Sentinel-1 and 14 Sentinel-2 images of Arctic glaciers, with one optical image per glacier in the dataset. Additionally, we introduce a novel hybrid model architecture that combines a Swin Transformer encoder with a residual Convolutional Neural Network (CNN) decoder. When pretrained on SSL4SAR, this model achieves a mean distance error of 293 m on the "CAlving Fronts and where to Find thEm" (CaFFe) benchmark dataset, outperforming the prior best model by 67 m. Evaluating an ensemble of the proposed model on a multi-annotator study of the benchmark dataset reveals a mean distance error of 75 m, approaching the human performance of 38 m. This advancement enables precise monitoring of seasonal changes in glacier calving fronts.

Mathias Seuret, Janne van der Loop, Nikolaus Weichselbaumer et al.

In this paper, we investigate the usage of fine-grained font recognition on OCR for books printed from the 15th to the 18th century. We used a newly created dataset for OCR of early printed books for which fonts are labeled with bounding boxes. We know not only the font group used for each character, but the locations of font changes as well. In books of this period, we frequently find font group changes mid-line or even mid-word that indicate changes in language. We consider 8 different font groups present in our corpus and investigate 13 different subsets: the whole dataset and text lines with a single font, multiple fonts, Roman fonts, Gothic fonts, and each of the considered fonts, respectively. We show that OCR performance is strongly impacted by font style and that selecting fine-tuned models with font group recognition has a very positive impact on the results. Moreover, we developed a system using local font group recognition in order to combine the output of multiple font recognition models, and show that while slower, this approach performs better not only on text lines composed of multiple fonts but on the ones containing a single font only as well.

Aline Sindel, Abner Hernandez, Seung Hee Yang et al.

With the increasing number of online learning material in the web, search for specific content in lecture videos can be time consuming. Therefore, automatic slide extraction from the lecture videos can be helpful to give a brief overview of the main content and to support the students in their studies. For this task, we propose a deep learning method to detect slide transitions in lectures videos. We first process each frame of the video by a heuristic-based approach using a 2-D convolutional neural network to predict transition candidates. Then, we increase the complexity by employing two 3-D convolutional neural networks to refine the transition candidates. Evaluation results demonstrate the effectiveness of our method in finding slide transitions.

Aline Sindel, Bettina Hohberger, Sebastian Fassihi Dehcordi et al.

Ophthalmological imaging utilizes different imaging systems, such as color fundus, infrared, fluorescein angiography, optical coherence tomography (OCT) or OCT angiography. Multiple images with different modalities or acquisition times are often analyzed for the diagnosis of retinal diseases. Automatically aligning the vessel structures in the images by means of multi-modal registration can support the ophthalmologists in their work. Our method uses a convolutional neural network to extract features of the vessel structure in multi-modal retinal images. We jointly train a keypoint detection and description network on small patches using a classification and a cross-modal descriptor loss function and apply the network to the full image size in the test phase. Our method demonstrates the best registration performance on our and a public multi-modal dataset in comparison to competing methods.

Sonja Kunzmann, Christian Marzahl, Felix Denzinger et al.

Annotating data, especially in the medical domain, requires expert knowledge and a lot of effort. This limits the amount and/or usefulness of available medical data sets for experimentation. Therefore, developing strategies to increase the number of annotations while lowering the needed domain knowledge is of interest. A possible strategy is the use of gamification, i.e. transforming the annotation task into a game. We propose an approach to gamify the task of annotating lung fluid cells from pathological whole slide images (WSIs). As the domain is unknown to non-expert annotators, we transform images of cells to the domain of flower images using a CycleGAN architecture. In this more assessable domain, non-expert annotators can be (t)asked to annotate different kinds of flowers in a playful setting. In order to provide a proof of concept, this work shows that the domain transfer is possible by evaluating an image classification network trained on real cell images and tested on the cell images generated by the CycleGAN network (reconstructed cell images) as well as real cell images. The classification network reaches an average accuracy of 94.73 % on the original lung fluid cells and 95.25 % on the transformed lung fluid cells, respectively. Our study lays the foundation for future research on gamification using CycleGANs.

Mathis Hoffmann, Johannes Hepp, Bernd Doll et al.

The individual causes for power loss of photovoltaic modules are investigated for quite some time. Recently, it has been shown that the power loss of a module is, for example, related to the fraction of inactive areas. While these areas can be easily identified from electroluminescense (EL) images, this is much harder for photoluminescence (PL) images. With this work, we close the gap between power regression from EL and PL images. We apply a deep convolutional neural network to predict the module power from PL images with a mean absolute error (MAE) of 4.4% or 11.7WP. Furthermore, we depict that regression maps computed from the embeddings of the trained network can be used to compute the localized power loss. Finally, we show that these regression maps can be used to identify inactive regions in PL images as well.

Alexander Mattick, Martin Mayr, Mathias Seuret et al.

As of recent generative adversarial networks have allowed for big leaps in the realism of generated images in diverse domains, not the least of which being handwritten text generation. The generation of realistic-looking hand-written text is important because it can be used for data augmentation in handwritten text recognition (HTR) systems or human-computer interaction. We propose SmartPatch, a new technique increasing the performance of current state-of-the-art methods by augmenting the training feedback with a tailored solution to mitigate pen-level artifacts. We combine the well-known patch loss with information gathered from the parallel trained handwritten text recognition system and the separate characters of the word. This leads to a more enhanced local discriminator and results in more realistic and higher-quality generated handwritten words.

Soroosh Tayebi Arasteh, Mehrpad Monajem, Vincent Christlein et al.

Twitter sentiment analysis, which often focuses on predicting the polarity of tweets, has attracted increasing attention over the last years, in particular with the rise of deep learning (DL). In this paper, we propose a new task: predicting the predominant sentiment among (first-order) replies to a given tweet. Therefore, we created RETWEET, a large dataset of tweets and replies manually annotated with sentiment labels. As a strong baseline, we propose a two-stage DL-based method: first, we create automatically labeled training data by applying a standard sentiment classifier to tweet replies and aggregating its predictions for each original tweet; our rationale is that individual errors made by the classifier are likely to cancel out in the aggregation step. Second, we use the automatically labeled data for supervised training of a neural network to predict reply sentiment from the original tweets. The resulting classifier is evaluated on the new RETWEET dataset, showing promising results, especially considering that it has been trained without any manually labeled data. Both the dataset and the baseline implementation are publicly available.

Kai Packhäuser, Sebastian Gündel, Nicolas Münster et al.

With the rise and ever-increasing potential of deep learning techniques in recent years, publicly available medical datasets became a key factor to enable reproducible development of diagnostic algorithms in the medical domain. Medical data contains sensitive patient-related information and is therefore usually anonymized by removing patient identifiers, e.g., patient names before publication. To the best of our knowledge, we are the first to show that a well-trained deep learning system is able to recover the patient identity from chest X-ray data. We demonstrate this using the publicly available large-scale ChestX-ray14 dataset, a collection of 112,120 frontal-view chest X-ray images from 30,805 unique patients. Our verification system is able to identify whether two frontal chest X-ray images are from the same person with an AUC of 0.9940 and a classification accuracy of 95.55%. We further highlight that the proposed system is able to reveal the same person even ten and more years after the initial scan. When pursuing a retrieval approach, we observe an mAP@R of 0.9748 and a precision@1 of 0.9963. Furthermore, we achieve an AUC of up to 0.9870 and a precision@1 of up to 0.9444 when evaluating our trained networks on external datasets such as CheXpert and the COVID-19 Image Data Collection. Based on this high identification rate, a potential attacker may leak patient-related information and additionally cross-reference images to obtain more information. Thus, there is a great risk of sensitive content falling into unauthorized hands or being disseminated against the will of the concerned patients. Especially during the COVID-19 pandemic, numerous chest X-ray datasets have been published to advance research. Therefore, such data may be vulnerable to potential attacks by deep learning-based re-identification algorithms.

Amirabbas Davari, Christoph Baller, Thorsten Seehaus et al.

Glacier calving front position (CFP) is an important glaciological variable. Traditionally, delineating the CFPs has been carried out manually, which was subjective, tedious and expensive. Automating this process is crucial for continuously monitoring the evolution and status of glaciers. Recently, deep learning approaches have been investigated for this application. However, the current methods get challenged by a severe class-imbalance problem. In this work, we propose to mitigate the class-imbalance between the calving front class and the non-calving front class by reformulating the segmentation problem into a pixel-wise regression task. A Convolutional Neural Network gets optimized to predict the distance values to the glacier front for each pixel in the image. The resulting distance map localizes the CFP and is further post-processed to extract the calving front line. We propose three post-processing methods, one method based on statistical thresholding, a second method based on conditional random fields (CRF), and finally the use of a second U-Net. The experimental results confirm that our approach significantly outperforms the state-of-the-art methods and produces accurate delineation. The Second U-Net obtains the best performance results, resulting in an average improvement of about 21% dice coefficient enhancement.

Amirabbas Davari, Saahil Islam, Thorsten Seehaus et al.

The vast majority of the outlet glaciers and ice streams of the polar ice sheets end in the ocean. Ice mass loss via calving of the glaciers into the ocean has increased over the last few decades. Information on the temporal variability of the calving front position provides fundamental information on the state of the glacier and ice stream, which can be exploited as calibration and validation data to enhance ice dynamics modeling. To identify the calving front position automatically, deep neural network-based semantic segmentation pipelines can be used to delineate the acquired SAR imagery. However, the extreme class imbalance is highly challenging for the accurate calving front segmentation in these images. Therefore, we propose the use of the Mathews correlation coefficient (MCC) as an early stopping criterion because of its symmetrical properties and its invariance towards class imbalance. Moreover, we propose an improvement to the distance map-based binary cross-entropy (BCE) loss function. The distance map adds context to the loss function about the important regions for segmentation and helps accounting for the imbalanced data. Using Mathews correlation coefficient as early stopping demonstrates an average 15% dice coefficient improvement compared to the commonly used BCE. The modified distance map loss further improves the segmentation performance by another 2%. These results are encouraging as they support the effectiveness of the proposed methods for segmentation problems suffering from extreme class imbalances.

Rosanna Dietrich-Sussner, Amirabbas Davari, Thorsten Seehaus et al.

Supervised machine learning requires a large amount of labeled data to achieve proper test results. However, generating accurately labeled segmentation maps on remote sensing imagery, including images from synthetic aperture radar (SAR), is tedious and highly subjective. In this work, we propose to alleviate the issue of limited training data by generating synthetic SAR images with the pix2pix algorithm. This algorithm uses conditional Generative Adversarial Networks (cGANs) to generate an artificial image while preserving the structure of the input. In our case, the input is a segmentation mask, from which a corresponding synthetic SAR image is generated. We present different models, perform a comparative study and demonstrate that this approach synthesizes convincing glaciers in SAR images with promising qualitative and quantitative results.

Andreas Hartmann, Amirabbas Davari, Thorsten Seehaus et al.

Fluctuations of the glacier calving front have an important influence over the ice flow of whole glacier systems. It is therefore important to precisely monitor the position of the calving front. However, the manual delineation of SAR images is a difficult, laborious and subjective task. Convolutional neural networks have previously shown promising results in automating the glacier segmentation in SAR images, making them desirable for further exploration of their possibilities. In this work, we propose to compute uncertainty and use it in an Uncertainty Optimization regime as a novel two-stage process. By using dropout as a random sampling layer in a U-Net architecture, we create a probabilistic Bayesian Neural Network. With several forward passes, we create a sampling distribution, which can estimate the model uncertainty for each pixel in the segmentation mask. The additional uncertainty map information can serve as a guideline for the experts in the manual annotation of the data. Furthermore, feeding the uncertainty map to the network leads to 95.24% Dice similarity, which is an overall improvement in the segmentation performance compared to the state-of-the-art deterministic U-Net-based glacier segmentation pipelines.

Michael Holzmann, Amirabbas Davari, Thorsten Seehaus et al.

An essential climate variable to determine the tidewater glacier status is the location of the calving front position and the separation of seasonal variability from long-term trends. Previous studies have proposed deep learning-based methods to semi-automatically delineate the calving fronts of tidewater glaciers. They used U-Net to segment the ice and non-ice regions and extracted the calving fronts in a post-processing step. In this work, we show a method to segment the glacier calving fronts from SAR images in an end-to-end fashion using Attention U-Net. The main objective is to investigate the attention mechanism in this application. Adding attention modules to the state-of-the-art U-Net network lets us analyze the learning process by extracting its attention maps. We use these maps as a tool to search for proper hyperparameters and loss functions in order to generate higher qualitative results. Our proposed attention U-Net performs comparably to the standard U-Net while providing additional insight into those regions on which the network learned to focus more. In the best case, the attention U-Net achieves a 1.5% better Dice score compared to the canonical U-Net with a glacier front line prediction certainty of up to 237.12 meters.

Prathmesh Madhu, Angel Villar-Corrales, Ronak Kosti et al.

Human pose estimation (HPE) is a central part of understanding the visual narration and body movements of characters depicted in artwork collections, such as Greek vase paintings. Unfortunately, existing HPE methods do not generalise well across domains resulting in poorly recognized poses. Therefore, we propose a two step approach: (1) adapting a dataset of natural images of known person and pose annotations to the style of Greek vase paintings by means of image style-transfer. We introduce a perceptually-grounded style transfer training to enforce perceptual consistency. Then, we fine-tune the base model with this newly created dataset. We show that using style-transfer learning significantly improves the SOTA performance on unlabelled data by more than 6% mean average precision (mAP) as well as mean average recall (mAR). (2) To improve the already strong results further, we created a small dataset (ClassArch) consisting of ancient Greek vase paintings from the 6-5th century BCE with person and pose annotations. We show that fine-tuning on this data with a style-transferred model improves the performance further. In a thorough ablation study, we give a targeted analysis of the influence of style intensities, revealing that the model learns generic domain styles. Additionally, we provide a pose-based image retrieval to demonstrate the effectiveness of our method.

Mathis Hoffmann, Thomas Köhler, Bernd Doll et al.

Visual inspection of solar modules is an important monitoring facility in photovoltaic power plants. Since a single measurement of fast CMOS sensors is limited in spatial resolution and often not sufficient to reliably detect small defects, we apply multi-frame super-resolution (MFSR) to a sequence of low resolution measurements. In addition, the rectification and removal of lens distortion simplifies subsequent analysis. Therefore, we propose to fuse this pre-processing with standard MFSR algorithms. This is advantageous, because we omit a separate processing step, the motion estimation becomes more stable and the spacing of high-resolution (HR) pixels on the rectified module image becomes uniform w. r. t. the module plane, regardless of perspective distortion. We present a comprehensive user study showing that MFSR is beneficial for defect recognition by human experts and that the proposed method performs better than the state of the art. Furthermore, we apply automated crack segmentation and show that the proposed method performs 3x better than bicubic upsampling and 2x better than the state of the art for automated inspection.

Mathias Seuret, Anguelos Nicolaou, Dominique Stutzmann et al.

This competition succeeds upon a line of competitions for writer and style analysis of historical document images. In particular, we investigate the performance of large-scale retrieval of historical document fragments in terms of style and writer identification. The analysis of historic fragments is a difficult challenge commonly solved by trained humanists. In comparison to previous competitions, we make the results more meaningful by addressing the issue of sample granularity and moving from writer to page fragment retrieval. The two approaches, style and author identification, provide information on what kind of information each method makes better use of and indirectly contribute to the interpretability of the participating method. Therefore, we created a large dataset consisting of more than 120 000 fragments. Although the most teams submitted methods based on convolutional neural networks, the winning entry achieves an mAP below 40%.

Mathis Hoffmann, Claudia Buerhop-Lutz, Luca Reeb et al.

Automated inspection plays an important role in monitoring large-scale photovoltaic power plants. Commonly, electroluminescense measurements are used to identify various types of defects on solar modules but have not been used to determine the power of a module. However, knowledge of the power at maximum power point is important as well, since drops in the power of a single module can affect the performance of an entire string. By now, this is commonly determined by measurements that require to discontact or even dismount the module, rendering a regular inspection of individual modules infeasible. In this work, we bridge the gap between electroluminescense measurements and the power determination of a module. We compile a large dataset of 719 electroluminescense measurementsof modules at various stages of degradation, especially cell cracks and fractures, and the corresponding power at maximum power point. Here,we focus on inactive regions and cracks as the predominant type of defect. We set up a baseline regression model to predict the power from electroluminescense measurements with a mean absolute error of 9.0+/-3.7$W_P$ (4.0+/-8.4%). Then, we show that deep-learning can be used to train a model that performs significantly better (7.3+/-2.7$W_P$ or 3.2+/-6.5%) and propose a variant of class activation maps to obtain the per cell power loss, as predicted by the model. With this work, we aim to open a new research topic. Therefore, we publicly release the dataset, the code and trained models to empower other researchers to compare against our results. Finally, we present a thorough evaluation of certain boundary conditions like the dataset size and an automated preprocessing pipeline for on-site measurements showing multiple modules at once.

Martin Leipert, Georg Vogeler, Mathias Seuret et al.

Notarial instruments are a category of documents. A notarial instrument can be distinguished from other documents by its notary sign, a prominent symbol in the certificate, which also allows to identify the document's issuer. Naturally, notarial instruments are underrepresented in regard to other documents. This makes a classification difficult because class imbalance in training data worsens the performance of Convolutional Neural Networks. In this work, we evaluate different countermeasures for this problem. They are applied to a binary classification and a segmentation task on a collection of medieval documents. In classification, notarial instruments are distinguished from other documents, while the notary sign is separated from the certificate in the segmentation task. We evaluate different techniques, such as data augmentation, under- and oversampling, as well as regularizing with focal loss. The combination of random minority oversampling and data augmentation leads to the best performance. In segmentation, we evaluate three loss-functions and their combinations, where only class-weighted dice loss was able to segment the notary sign sufficiently.

Vincent Christlein, Nikolaus Weichselbaumer, Saskia Limbach et al.

The type used to print an early modern book can give scholars valuable information about the time and place of its production as well as its producer. Recognizing such type is currently done manually using both the character shapes of `M' or `Qu' and the size of the total type to look it up in a large reference work. This is a reliable method, but it is also slow and requires specific skills. We investigate the performance of type classification and type retrieval using a newly created dataset consisting of easy and difficult types used in early printed books. For type classification, we rely on a deep Convolutional Neural Network (CNN) originally used for font-group classification while we use a common writer identification method for the retrieval case. We show that in both scenarios, easy types can be classified/retrieved with a high accuracy while difficult cases are indeed difficult.

Simon Jordan, Mathias Seuret, Pavel Král et al.

Automatic writer identification is a common problem in document analysis. State-of-the-art methods typically focus on the feature extraction step with traditional or deep-learning-based techniques. In retrieval problems, re-ranking is a commonly used technique to improve the results. Re-ranking refines an initial ranking result by using the knowledge contained in the ranked result, e. g., by exploiting nearest neighbor relations. To the best of our knowledge, re-ranking has not been used for writer identification/retrieval. A possible reason might be that publicly available benchmark datasets contain only few samples per writer which makes a re-ranking less promising. We show that a re-ranking step based on k-reciprocal nearest neighbor relationships is advantageous for writer identification, even if only a few samples per writer are available. We use these reciprocal relationships in two ways: encode them into new vectors, as originally proposed, or integrate them in terms of query-expansion. We show that both techniques outperform the baseline results in terms of mAP on three writer identification datasets.

Martin Mayr, Martin Stumpf, Anguelos Nicolaou et al.

Most people think that their handwriting is unique and cannot be imitated by machines, especially not using completely new content. Current cursive handwriting synthesis is visually limited or needs user interaction. We show that subdividing the process into smaller subtasks makes it possible to imitate someone's handwriting with a high chance to be visually indistinguishable for humans. Therefore, a given handwritten sample will be used as the target style. This sample is transferred to an online sequence. Then, a method for online handwriting synthesis is used to produce a new realistic-looking text primed with the online input sequence. This new text is then rendered and style-adapted to the input pen. We show the effectiveness of the pipeline by generating in- and out-of-vocabulary handwritten samples that are validated in a comprehensive user study. Additionally, we show that also a typical writer identification system can partially be fooled by the created fake handwritings.