Abstract: This four-volume set of LNCS 12821, LNCS 12822, LNCS 12823 and LNCS 12824, constitutes the refereed proceedings of the 16th International Conference on Document Analysis and Recognition, ICDAR 2021, held in Lausanne, Switzerland in September 2021. The 182 full papers were carefully reviewed and selected from 340 submissions, and are presented with 13 competition reports.
The papers are organized into the following topical sections: historical document analysis, document analysis systems, handwriting recognition, scene text detection and recognition, document image processing, natural language processing (NLP) for document understanding, and graphics, diagram and math recognition.

Abstract: The study of mental workload becomes essential for human work efficiency, health conditions and to avoid accidents, since workload compromises both performance and awareness. Although workload has been widely studied using several physiological measures, minimising the sensor network as much as possible remains both a challenge and a requirement.
Electroencephalogram (EEG) signals have shown a high correlation to specific cognitive and mental states like workload. However, there is not enough evidence in the literature to validate how well models generalize in case of new subjects performing tasks of a workload similar to the ones included during model’s training.
In this paper we propose a binary neural network to classify EEG features across different mental workloads. Two workloads, low and medium, are induced using two variants of the N-Back Test. The proposed model was validated in a dataset collected from 16 subjects and shown a high level of generalization capability: model reported an average recall of 81.81% in a leave-one-out subject evaluation.

Abstract: People suffering Bipolar Disorder (BD) experiment changes in mood status having depressive or manic episodes with normal periods in the middle. BD is a chronic disease with a high level of non-adherence to medication that needs a continuous monitoring of patients to detect when they relapse in an episode, so that physicians can take care of them. Here we present MoodRecord, an easy-to-use, non-intrusive, multilingual, robust and scalable platform suitable for home monitoring patients with BD, that allows physicians and relatives to track the patient state and get alarms when abnormalities occur.
MoodRecord takes advantage of the capabilities of smartphones as a communication and recording device to do a continuous monitoring of patients. It automatically records user activity, and asks the user to answer some questions or to record himself in video, according to a predefined plan designed by physicians. The video is analysed, recognising the mood status from images and bipolar assessment scores are extracted from speech parameters. The data obtained from the different sources are merged periodically to observe if a relapse may start and if so, raise the corresponding alarm. The application got a positive evaluation in a pilot with users from three different countries. During the pilot, the predictions of the voice and image modules showed a coherent correlation with the diagnosis performed by clinicians.

Abstract: This work presents Satellite Style and Structure Generative Adversarial Network (SSGAN), a generative model of high resolution satellite imagery to support image segmentation. Based on spatially adaptive denormalization modules (SPADE) that modulate the activations with respect to segmentation map structure, in addition to global descriptor vectors that capture the semantic information in a vector with respect to Open Street Maps (OSM) classes, this model is able to produce
consistent aerial imagery. By decoupling the generation of aerial images into a structure map and a carefully defined style vector, we were able to improve the realism and geodiversity of the synthesis with respect to the state-of-the-art baseline. Therefore, the proposed model allows us to control the generation not only with respect to the desired structure, but also with respect to a geographic area.

Abstract: Active learning is a paradigm aimed at reducing the annotation effort by training the model on actively selected informative and/or representative samples. Another paradigm to reduce the annotation effort is self-training that learns from a large amount of unlabeled data in an unsupervised way and fine-tunes on few labeled samples. Recent developments in self-training have achieved very impressive results rivaling supervised learning on some datasets. The current work focuses on whether the two paradigms can benefit from each other. We studied object recognition datasets including CIFAR10, CIFAR100 and Tiny ImageNet with several labeling budgets for the evaluations. Our experiments reveal that self-training is remarkably more efficient than active learning at reducing the labeling effort, that for a low labeling budget, active learning offers no benefit to self-training, and finally that the combination of active learning and self-training is fruitful when the labeling budget is high. The performance gap between active learning trained either with self-training or from scratch diminishes as we approach to the point where almost half of the dataset is labeled.

Abstract: Deep convolutional neural networks (CNNs) have achieved superior performance in many visual recognition applications, such as image classification, detection and segmentation. Training deep CNNs requires huge amounts of labeled data, which is expensive and labor intensive to collect. Active learning is a paradigm aimed at reducing the annotation effort by training the model on actively selected
informative and/or representative samples. In this thesis we study several aspects of active learning including video object detection for autonomous driving systems, image classification on balanced and imbalanced datasets and the incorporation of self-supervised learning in active learning. We briefly describe our approach in each of these areas to reduce the labeling effort.
In chapter two we introduce a novel active learning approach for object detection in videos by exploiting temporal coherence. Our criterion is based on the estimated number of errors in terms of false positives and false negatives. Additionally, we introduce a synthetic video dataset, called SYNTHIA-AL, specially designed to evaluate active
learning for video object detection in road scenes. Finally, we show that our
approach outperforms active learning baselines tested on two outdoor datasets.
In the next chapter we address the well-known problem of over confidence in the neural networks. As an alternative to network confidence, we propose a new informativeness-based active learning method that captures the learning dynamics of neural network with a metric called label-dispersion. This metric is low when the network consistently assigns the same label to the sample during the course of training and high when the assigned label changes frequently. We show that label-dispersion is a promising predictor of the uncertainty of the network, and show on two benchmark datasets that an active learning algorithm based on label-dispersion obtains excellent results.
In chapter four, we tackle the problem of sampling bias in active learning methods on imbalanced datasets. Active learning is generally studied on balanced datasets where an equal amount of images per class is available. However, real-world datasets suffer from severe imbalanced classes, the so called longtail distribution. We argue that this further complicates the active learning process, since the imbalanced data pool can result in suboptimal classifiers. To address this problem in the context of active learning, we propose a general optimization framework that explicitly takes class-balancing into account. Results on three datasets show that the method is general (it can be combined with most existing active learning algorithms) and can be effectively applied to boost the performance of both informative and representative-based active learning methods. In addition, we show that also on balanced datasets our method generally results in a performance gain.
Another paradigm to reduce the annotation effort is self-training that learns from a large amount of unlabeled data in an unsupervised way and fine-tunes on few labeled samples. Recent advancements in self-training have achieved very impressive results rivaling supervised learning on some datasets. In the last chapter we focus on whether active learning and self supervised learning can benefit from each other.
We study object recognition datasets with several labeling budgets for the evaluations. Our experiments reveal that self-training is remarkably more efficient than active learning at reducing the labeling effort, that for a low labeling budget, active learning offers no benefit to self-training, and finally that the combination of active learning and self-training is fruitful when the labeling budget is high.

Abstract: We present a methodology to retrieve analytical surfaces parametrized as a neural network. Previous works on 3D reconstruction yield point clouds, voxelized objects or meshes. Instead, our approach yields 2-manifolds in the euclidean space through deep learning. To this end, we implement a novel formulation for fully connected layers as parametrized manifolds that allows continuous predictions with differential geometry. Based on this property we propose a novel smoothness loss. Results on CLOTH3D++ dataset show the possibility to infer different topologies and the benefits of the smoothness term based on differential geometry.