Abstract: Rain removal plays an important role in the restoration of degraded images. Recently, CNN-based methods have achieved remarkable success. However, these approaches neglect that the appearance of real-world rain is often accompanied by low light conditions, which will further degrade the image quality, thereby hindering the restoration mission. Therefore, it is very indispensable to jointly remove the rain and enhance illumination for real-world rain image restoration. To this end, we proposed a novel spatially-adaptive network, dubbed SANet, which can remove the rain and enhance illumination in one go with the guidance of degradation mask. Meanwhile, to fully utilize negative samples, a contrastive loss is proposed to preserve more natural textures and consistent illumination. In addition, we present a new synthetic dataset, named DarkRain, to boost the development of rain image restoration algorithms in practical scenarios. DarkRain not only contains different degrees of rain, but also considers different lighting conditions, and more realistically simulates real-world rainfall scenarios. SANet is extensively evaluated on the proposed dataset and attains new state-of-the-art performance against other combining methods. Moreover, after a simple transformation, our SANet surpasses existing the state-of-the-art algorithms in both rain removal and low-light image enhancement.

Abstract: Edge detection is the basis of many computer vision applications. State of the art predominantly relies on deep learning with two decisive factors: dataset content and network architecture. Most of the publicly available datasets are not curated for edge detection tasks. Here, we address this limitation. First, we argue that edges, contours and boundaries, despite their overlaps, are three distinct visual features requiring separate benchmark datasets. To this end, we present a new dataset of edges. Second, we propose a novel architecture, termed Dense Extreme Inception Network for Edge Detection (DexiNed), that can be trained from scratch without any pre-trained weights. DexiNed outperforms other algorithms in the presented dataset. It also generalizes well to other datasets without any fine-tuning. The higher quality of DexiNed is also perceptually evident thanks to the sharper and finer edges it outputs.

Abstract: Document Visual Question Answering (DocVQA) refers to the task of answering questions from document images. Existing work on DocVQA only considers single-page documents. However, in real scenarios documents are mostly composed of multiple pages that should be processed altogether. In this work we extend DocVQA to the multi-page scenario. For that, we first create a new dataset, MP-DocVQA, where questions are posed over multi-page documents instead of single pages. Second, we propose a new hierarchical method, Hi-VT5, based on the T5 architecture, that overcomes the limitations of current methods to process long multi-page documents. The proposed method is based on a hierarchical transformer architecture where the encoder summarizes the most relevant information of every page and then, the decoder takes this summarized information to generate the final answer. Through extensive experimentation, we demonstrate that our method is able, in a single stage, to answer the questions and provide the page that contains the relevant information to find the answer, which can be used as a kind of explainability measure.

Abstract: Hand sign language recognition from video is a challenging research area in computer vision, which performance is affected by hand occlusion, fast hand movement, illumination changes, or background complexity, just to mention a few. In recent years, deep learning approaches have achieved state-of-the-art results in the field, though previous challenges are not completely solved. In this work, we propose a novel deep learning-based pipeline architecture for efficient automatic hand sign language recognition using Single Shot Detector (SSD), 2D Convolutional Neural Network (2DCNN), 3D Convolutional Neural Network (3DCNN), and Long Short-Term Memory (LSTM) from RGB input videos. We use a CNN-based model which estimates the 3D hand keypoints from 2D input frames. After that, we connect these estimated keypoints to build the hand skeleton by using midpoint algorithm. In order to obtain a more discriminative representation of hands, we project 3D hand skeleton into three views surface images. We further employ the heatmap image of detected keypoints as input for refinement in a stacked fashion. We apply 3DCNNs on the stacked features of hand, including pixel level, multi-view hand skeleton, and heatmap features, to extract discriminant local spatio-temporal features from these stacked inputs. The outputs of the 3DCNNs are fused and fed to a LSTM to model long-term dynamics of hand sign gestures. Analyzing 2DCNN vs. 3DCNN using different number of stacked inputs into the network, we demonstrate that 3DCNN better capture spatio-temporal dynamics of hands. To the best of our knowledge, this is the first time that this multi-modal and multi-view set of hand skeleton features are applied for hand sign language recognition. Furthermore, we present a new large-scale hand sign language dataset, namely RKS-PERSIANSIGN, including 10′000 RGB videos of 100 Persian sign words. Evaluation results of the proposed model on three datasets, NYU, First-Person, and RKS-PERSIANSIGN, indicate that our model outperforms state-of-the-art models in hand sign language recognition, hand pose estimation, and hand action recognition.

Abstract: Nutrition and social interactions are both key aspects of the daily lives of humans. In this work, we propose a system to evaluate the influence of social interaction in the nutritional habits of a person from a first-person perspective. In order to detect the routine of an individual, we construct a nutritional behaviour pattern discovery model, which outputs routines over a number of days. Our method evaluates similarity of routines with respect to visited food-related scenes over the collected days, making use of Dynamic Time Warping, as well as considering social engagement and its correlation with food-related activities. The nutritional and social descriptors of the collected days are evaluated and encoded using an LSTM Autoencoder. Later, the obtained latent space is clustered to find similar days unaffected by outliers using the Isolation Forest method. Moreover, we introduce a new score metric to evaluate the performance of the proposed algorithm. We validate our method on 104 days and more than 100 k egocentric images gathered by 7 users. Several different visualizations are evaluated for the understanding of the findings. Our results demonstrate good performance and applicability of our proposed model for social-related nutritional behaviour understanding. At the end, relevant applications of the model are discussed by analysing the discovered routine of particular individuals.

Abstract: The analysis of urinary stones is mandatory for the best management of the disease after the stone passage in order to prevent further stone episodes. Thus the use of an appropriate methodology for an individualized stone analysis becomes a key factor for giving the patient the most suitable treatment. A recently developed hyperspectral imaging methodology, based on pixel-to-pixel analysis of near-infrared spectral images, is compared to the reference technique in stone analysis, infrared (IR) spectroscopy. The developed classification model yields >90% correct classification rate when compared to IR and is able to precisely locate stone components within the structure of the stone with a 15 µm resolution. Due to the little sample pretreatment, low analysis time, good performance of the model, and the automation of the measurements, they become analyst independent; this methodology can be considered to become a routine analysis for clinical laboratories.

Abstract: The segmentation of visible electromagnetic radiation into chromatic categories by the human visual system has been extensively studied from a perceptual point of view, resulting in several colour appearance models. However, there is currently a void when it comes to relate these results to the physiological mechanisms that are known to shape the pre-cortical and cortical visual pathway. This work intends to begin to fill this void by proposing a new physiologically plausible model of colour categorization based on Neural Isoresponsive Colour Ellipsoids (NICE) in the cone-contrast space defined by the main directions of the visual signals entering the visual cortex. The model was adjusted to fit psychophysical measures that concentrate on the categorical boundaries and are consistent with the ellipsoidal isoresponse surfaces of visual cortical neurons. By revealing the shape of such categorical colour regions, our measures allow for a more precise and parsimonious description, connecting well-known early visual processing mechanisms to the less understood phenomenon of colour categorization. To test the feasibility of our method we applied it to exemplary images and a popular ground-truth chart obtaining labelling results that are better than those of current state-of-the-art algorithms.

Abstract: A main bottleneck in bronchoscopic biopsy sampling is to efficiently reach the lesion navigating across bronchial levels. Any guidance system should be able to localize the scope position during the intervention with minimal costs and alteration of clinical protocols. With the final goal of an affordable image-based guidance, this work presents a novel strategy to extract and codify the anatomical structure of bronchi, as well as, the scope navigation path from videobronchoscopy. Experiments using interventional data show that our method accurately identifies the bronchial structure. Meanwhile, experiments using simulated data verify that the extracted navigation path matches the 3D route.