Josep Llados, Daniel Lopresti, & Seiichi Uchida (Eds.). (2021). 16th International Conference, 2021, Proceedings, Part II (Vol. 12822). LNCS. Springer Cham.
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: document analysis for literature search, document summarization and translation, multimedia document analysis, mobile text recognition, document analysis for social good, indexing and retrieval of documents, physical and logical layout analysis, recognition of tables and formulas, and natural language processing (NLP) for document understanding.
|
Ruben Tito, Dimosthenis Karatzas, & Ernest Valveny. (2021). Document Collection Visual Question Answering. In 16th International Conference on Document Analysis and Recognition (Vol. 12822, pp. 778–792). LNCS.
Abstract: Current tasks and methods in Document Understanding aims to process documents as single elements. However, documents are usually organized in collections (historical records, purchase invoices), that provide context useful for their interpretation. To address this problem, we introduce Document Collection Visual Question Answering (DocCVQA) a new dataset and related task, where questions are posed over a whole collection of document images and the goal is not only to provide the answer to the given question, but also to retrieve the set of documents that contain the information needed to infer the answer. Along with the dataset we propose a new evaluation metric and baselines which provide further insights to the new dataset and task.
Keywords: Document collection; Visual Question Answering
|
Ruben Tito, Minesh Mathew, C.V. Jawahar, Ernest Valveny, & Dimosthenis Karatzas. (2021). ICDAR 2021 Competition on Document Visual Question Answering. In 16th International Conference on Document Analysis and Recognition (pp. 635–649).
Abstract: In this report we present results of the ICDAR 2021 edition of the Document Visual Question Challenges. This edition complements the previous tasks on Single Document VQA and Document Collection VQA with a newly introduced on Infographics VQA. Infographics VQA is based on a new dataset of more than 5, 000 infographics images and 30, 000 question-answer pairs. The winner methods have scored 0.6120 ANLS in Infographics VQA task, 0.7743 ANLSL in Document Collection VQA task and 0.8705 ANLS in Single Document VQA. We present a summary of the datasets used for each task, description of each of the submitted methods and the results and analysis of their performance. A summary of the progress made on Single Document VQA since the first edition of the DocVQA 2020 challenge is also presented.
|
Albert Suso, Pau Riba, Oriol Ramos Terrades, & Josep Llados. (2021). A Self-supervised Inverse Graphics Approach for Sketch Parametrization. In 16th International Conference on Document Analysis and Recognition (Vol. 12916, pp. 28–42). LNCS.
Abstract: The study of neural generative models of handwritten text and human sketches is a hot topic in the computer vision field. The landmark SketchRNN provided a breakthrough by sequentially generating sketches as a sequence of waypoints, and more recent articles have managed to generate fully vector sketches by coding the strokes as Bézier curves. However, the previous attempts with this approach need them all a ground truth consisting in the sequence of points that make up each stroke, which seriously limits the datasets the model is able to train in. In this work, we present a self-supervised end-to-end inverse graphics approach that learns to embed each image to its best fit of Bézier curves. The self-supervised nature of the training process allows us to train the model in a wider range of datasets, but also to perform better after-training predictions by applying an overfitting process on the input binary image. We report qualitative an quantitative evaluations on the MNIST and the Quick, Draw! datasets.
|
Reza Azad, Afshin Bozorgpour, Maryam Asadi-Aghbolaghi, Dorit Merhof, & Sergio Escalera. (2021). Deep Frequency Re-Calibration U-Net for Medical Image Segmentation. In IEEE/CVF International Conference on Computer Vision Workshops (pp. 3274–3283).
Abstract: We present a novel solution to the garment animation problem through deep learning. Our contribution allows animating any template outfit with arbitrary topology and geometric complexity. Recent works develop models for garment edition, resizing and animation at the same time by leveraging the support body model (encoding garments as body homotopies). This leads to complex engineering solutions that suffer from scalability, applicability and compatibility. By limiting our scope to garment animation only, we are able to propose a simple model that can animate any outfit, independently of its topology, vertex order or connectivity. Our proposed architecture maps outfits to animated 3D models into the standard format for 3D animation (blend weights and blend shapes matrices), automatically providing of compatibility with any graphics engine. We also propose a methodology to complement supervised learning with an unsupervised physically based learning that implicitly solves collisions and enhances cloth quality.
|
Ajian Liu, Chenxu Zhao, Zitong Yu, Anyang Su, Xing Liu, Zijian Kong, et al. (2021). 3D High-Fidelity Mask Face Presentation Attack Detection Challenge. In IEEE/CVF International Conference on Computer Vision Workshops (pp. 814–823).
Abstract: The threat of 3D mask to face recognition systems is increasing serious, and has been widely concerned by researchers. To facilitate the study of the algorithms, a large-scale High-Fidelity Mask dataset, namely CASIA-SURF HiFiMask (briefly HiFiMask) has been collected. Specifically, it consists of total amount of 54,600 videos which are recorded from 75 subjects with 225 realistic masks under 7 new kinds of sensors. Based on this dataset and Protocol 3 which evaluates both the discrimination and generalization ability of the algorithm under the open set scenarios, we organized a 3D High-Fidelity Mask Face Presentation Attack Detection Challenge to boost the research of 3D mask based attack detection. It attracted more than 200 teams for the development phase with a total of 18 teams qualifying for the final round. All the results were verified and re-ran by the organizing team, and the results were used for the final ranking. This paper presents an overview of the challenge, including the introduction of the dataset used, the definition of the protocol, the calculation of the evaluation criteria, and the summary and publication of the competition results. Finally, we focus on introducing and analyzing the top ranked algorithms, the conclusion summary, and the research ideas for mask attack detection provided by this competition.
|
Claudia Greco, Carmela Buono, Pau Buch-Cardona, Gennaro Cordasco, Sergio Escalera, Anna Esposito, et al. (2021). Emotional Features of Interactions With Empathic Agents. In IEEE/CVF International Conference on Computer Vision Workshops (pp. 2168–2176).
Abstract: The current study is part of the EMPATHIC project, whose aim is to develop an Empathic Virtual Coach (VC) capable of promoting healthy and independent aging. To this end, the VC needs to be capable of perceiving the emotional states of users and adjusting its behaviour during the interactions according to what the users are experiencing in terms of emotions and comfort. Thus, the present work focuses on some sessions where elderly users of three different countries interact with a simulated system. Audio and video information extracted from these sessions were examined by external observers to assess participants' emotional experience with the EMPATHIC-VC in terms of categorical and dimensional assessment of emotions. Analyses were conducted on the emotional labels assigned by the external observers while participants were engaged in two different scenarios: a generic one, where the interaction was carried out with no intention to discuss a specific topic, and a nutrition one, aimed to accomplish a conversation on users' nutritional habits. Results of analyses performed on both audio and video data revealed that the EMPATHIC coach did not elicit negative feelings in the users. Indeed, users from all countries have shown relaxed and positive behavior when interacting with the simulated VC during both scenarios. Overall, the EMPATHIC-VC was capable to offer an enjoyable experience without eliciting negative feelings in the users. This supports the hypothesis that an Empathic Virtual Coach capable of considering users' expectations and emotional states could support elderly people in daily life activities and help them to remain independent.
|
David Curto, Albert Clapes, Javier Selva, Sorina Smeureanu, Julio C. S. Jacques Junior, David Gallardo-Pujol, et al. (2021). Dyadformer: A Multi-Modal Transformer for Long-Range Modeling of Dyadic Interactions. In IEEE/CVF International Conference on Computer Vision Workshops (pp. 2177–2188).
Abstract: Personality computing has become an emerging topic in computer vision, due to the wide range of applications it can be used for. However, most works on the topic have focused on analyzing the individual, even when applied to interaction scenarios, and for short periods of time. To address these limitations, we present the Dyadformer, a novel multi-modal multi-subject Transformer architecture to model individual and interpersonal features in dyadic interactions using variable time windows, thus allowing the capture of long-term interdependencies. Our proposed cross-subject layer allows the network to explicitly model interactions among subjects through attentional operations. This proof-of-concept approach shows how multi-modality and joint modeling of both interactants for longer periods of time helps to predict individual attributes. With Dyadformer, we improve state-of-the-art self-reported personality inference results on individual subjects on the UDIVA v0.5 dataset.
|
Neelu Madan, Arya Farkhondeh, Kamal Nasrollahi, Sergio Escalera, & Thomas B. Moeslund. (2021). Temporal Cues From Socially Unacceptable Trajectories for Anomaly Detection. In IEEE/CVF International Conference on Computer Vision Workshops (pp. 2150–2158).
Abstract: State-of-the-Art (SoTA) deep learning-based approaches to detect anomalies in surveillance videos utilize limited temporal information, including basic information from motion, e.g., optical flow computed between consecutive frames. In this paper, we compliment the SoTA methods by including long-range dependencies from trajectories for anomaly detection. To achieve that, we first created trajectories by running a tracker on two SoTA datasets, namely Avenue and Shanghai-Tech. We propose a prediction-based anomaly detection method using trajectories based on Social GANs, also called in this paper as temporal-based anomaly detection. Then, we hypothesize that late fusion of the result of this temporal-based anomaly detection system with spatial-based anomaly detection systems produces SoTA results. We verify this hypothesis on two spatial-based anomaly detection systems. We show that both cases produce results better than baseline spatial-based systems, indicating the usefulness of the temporal information coming from the trajectories for anomaly detection. We observe that the proposed approach depicts the maximum improvement in micro-level Area-Under-the-Curve (AUC) by 4.1% on CUHK Avenue and 3.4% on Shanghai-Tech over one of the baseline method. We also show a high performance on cross-data evaluation, where we learn the weights to combine spatial and temporal information on Shanghai-Tech and perform evaluation on CUHK Avenue and vice-versa.
|
Javad Zolfaghari Bengar, Joost Van de Weijer, Bartlomiej Twardowski, & Bogdan Raducanu. (2021). Reducing Label Effort: Self- Supervised Meets Active Learning. In International Conference on Computer Vision Workshops (pp. 1631–1639).
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.
|
Shun Yao, Fei Yang, Yongmei Cheng, & Mikhail Mozerov. (2021). 3D Shapes Local Geometry Codes Learning with SDF. In International Conference on Computer Vision Workshops (pp. 2110–2117).
Abstract: A signed distance function (SDF) as the 3D shape description is one of the most effective approaches to represent 3D geometry for rendering and reconstruction. Our work is inspired by the state-of-the-art method DeepSDF [17] that learns and analyzes the 3D shape as the iso-surface of its shell and this method has shown promising results especially in the 3D shape reconstruction and compression domain. In this paper, we consider the degeneration problem of reconstruction coming from the capacity decrease of the DeepSDF model, which approximates the SDF with a neural network and a single latent code. We propose Local Geometry Code Learning (LGCL), a model that improves the original DeepSDF results by learning from a local shape geometry of the full 3D shape. We add an extra graph neural network to split the single transmittable latent code into a set of local latent codes distributed on the 3D shape. Mentioned latent codes are used to approximate the SDF in their local regions, which will alleviate the complexity of the approximation compared to the original DeepSDF. Furthermore, we introduce a new geometric loss function to facilitate the training of these local latent codes. Note that other local shape adjusting methods use the 3D voxel representation, which in turn is a problem highly difficult to solve or even is insolvable. In contrast, our architecture is based on graph processing implicitly and performs the learning regression process directly in the latent code space, thus make the proposed architecture more flexible and also simple for realization. Our experiments on 3D shape reconstruction demonstrate that our LGCL method can keep more details with a significantly smaller size of the SDF decoder and outperforms considerably the original DeepSDF method under the most important quantitative metrics.
|
Yaxing Wang, Hector Laria Mantecon, Joost Van de Weijer, Laura Lopez-Fuentes, & Bogdan Raducanu. (2021). TransferI2I: Transfer Learning for Image-to-Image Translation from Small Datasets. In 19th IEEE International Conference on Computer Vision (pp. 13990–13999).
Abstract: Image-to-image (I2I) translation has matured in recent years and is able to generate high-quality realistic images. However, despite current success, it still faces important challenges when applied to small domains. Existing methods use transfer learning for I2I translation, but they still require the learning of millions of parameters from scratch. This drawback severely limits its application on small domains. In this paper, we propose a new transfer learning for I2I translation (TransferI2I). We decouple our learning process into the image generation step and the I2I translation step. In the first step we propose two novel techniques: source-target initialization and self-initialization of the adaptor layer. The former finetunes the pretrained generative model (e.g., StyleGAN) on source and target data. The latter allows to initialize all non-pretrained network parameters without the need of any data. These techniques provide a better initialization for the I2I translation step. In addition, we introduce an auxiliary GAN that further facilitates the training of deep I2I systems even from small datasets. In extensive experiments on three datasets, (Animal faces, Birds, and Foods), we show that we outperform existing methods and that mFID improves on several datasets with over 25 points.
|
Hugo Bertiche, Meysam Madadi, Emilio Tylson, & Sergio Escalera. (2021). DeePSD: Automatic Deep Skinning And Pose Space Deformation For 3D Garment Animation. In 19th IEEE International Conference on Computer Vision (pp. 5471–5480).
Abstract: We present a novel solution to the garment animation problem through deep learning. Our contribution allows animating any template outfit with arbitrary topology and geometric complexity. Recent works develop models for garment edition, resizing and animation at the same time by leveraging the support body model (encoding garments as body homotopies). This leads to complex engineering solutions that suffer from scalability, applicability and compatibility. By limiting our scope to garment animation only, we are able to propose a simple model that can animate any outfit, independently of its topology, vertex order or connectivity. Our proposed architecture maps outfits to animated 3D models into the standard format for 3D animation (blend weights and blend shapes matrices), automatically providing of compatibility with any graphics engine. We also propose a methodology to complement supervised learning with an unsupervised physically based learning that implicitly solves collisions and enhances cloth quality.
|
Jialuo Chen, Mohamed Ali Souibgui, Alicia Fornes, & Beata Megyesi. (2021). Unsupervised Alphabet Matching in Historical Encrypted Manuscript Images. In 4th International Conference on Historical Cryptology (pp. 34–37).
Abstract: Historical ciphers contain a wide range ofsymbols from various symbol sets. Iden-tifying the cipher alphabet is a prerequi-site before decryption can take place andis a time-consuming process. In this workwe explore the use of image processing foridentifying the underlying alphabet in ci-pher images, and to compare alphabets be-tween ciphers. The experiments show thatciphers with similar alphabets can be suc-cessfully discovered through clustering.
|
Pau Torras, Mohamed Ali Souibgui, Jialuo Chen, & Alicia Fornes. (2021). A Transcription Is All You Need: Learning to Align through Attention. In 14th IAPR International Workshop on Graphics Recognition (Vol. 12916, 141–146). LNCS.
Abstract: Historical ciphered manuscripts are a type of document where graphical symbols are used to encrypt their content instead of regular text. Nowadays, expert transcriptions can be found in libraries alongside the corresponding manuscript images. However, those transcriptions are not aligned, so these are barely usable for training deep learning-based recognition methods. To solve this issue, we propose a method to align each symbol in the transcript of an image with its visual representation by using an attention-based Sequence to Sequence (Seq2Seq) model. The core idea is that, by learning to recognise symbols sequence within a cipher line image, the model also identifies their position implicitly through an attention mechanism. Thus, the resulting symbol segmentation can be later used for training algorithms. The experimental evaluation shows that this method is promising, especially taking into account the small size of the cipher dataset.
|
