|
|
Alvaro Peris, Marc Bolaños, Petia Radeva, & Francisco Casacuberta. (2016). Video Description Using Bidirectional Recurrent Neural Networks. In 25th International Conference on Artificial Neural Networks (Vol. 2, pp. 3–11).
Abstract: Although traditionally used in the machine translation field, the encoder-decoder framework has been recently applied for the generation of video and image descriptions. The combination of Convolutional and Recurrent Neural Networks in these models has proven to outperform the previous state of the art, obtaining more accurate video descriptions. In this work we propose pushing further this model by introducing two contributions into the encoding stage. First, producing richer image representations by combining object and location information from Convolutional Neural Networks and second, introducing Bidirectional Recurrent Neural Networks for capturing both forward and backward temporal relationships in the input frames.
Keywords: Video description; Neural Machine Translation; Birectional Recurrent Neural Networks; LSTM; Convolutional Neural Networks
|
|
|
|
Eloi Puertas, Miguel Angel Bautista, Daniel Sanchez, Sergio Escalera, & Oriol Pujol. (2014). Learning to Segment Humans by Stacking their Body Parts,. In ECCV Workshop on ChaLearn Looking at People (Vol. 8925, pp. 685–697). LNCS.
Abstract: Human segmentation in still images is a complex task due to the wide range of body poses and drastic changes in environmental conditions. Usually, human body segmentation is treated in a two-stage fashion. First, a human body part detection step is performed, and then, human part detections are used as prior knowledge to be optimized by segmentation strategies. In this paper, we present a two-stage scheme based on Multi-Scale Stacked Sequential Learning (MSSL). We define an extended feature set by stacking a multi-scale decomposition of body
part likelihood maps. These likelihood maps are obtained in a first stage
by means of a ECOC ensemble of soft body part detectors. In a second stage, contextual relations of part predictions are learnt by a binary classifier, obtaining an accurate body confidence map. The obtained confidence map is fed to a graph cut optimization procedure to obtain the final segmentation. Results show improved segmentation when MSSL is included in the human segmentation pipeline.
Keywords: Human body segmentation; Stacked Sequential Learning
|
|
|
|
Cristina Palmero, Albert Clapes, Chris Bahnsen, Andreas Møgelmose, Thomas B. Moeslund, & Sergio Escalera. (2016). Multi-modal RGB-Depth-Thermal Human Body Segmentation. IJCV - International Journal of Computer Vision, 118(2), 217–239.
Abstract: This work addresses the problem of human body segmentation from multi-modal visual cues as a first stage of automatic human behavior analysis. We propose a novel RGB–depth–thermal dataset along with a multi-modal segmentation baseline. The several modalities are registered using a calibration device and a registration algorithm. Our baseline extracts regions of interest using background subtraction, defines a partitioning of the foreground regions into cells, computes a set of image features on those cells using different state-of-the-art feature extractions, and models the distribution of the descriptors per cell using probabilistic models. A supervised learning algorithm then fuses the output likelihoods over cells in a stacked feature vector representation. The baseline, using Gaussian mixture models for the probabilistic modeling and Random Forest for the stacked learning, is superior to other state-of-the-art methods, obtaining an overlap above 75 % on the novel dataset when compared to the manually annotated ground-truth of human segmentations.
Keywords: Human body segmentation; RGB ; Depth Thermal
|
|
|
|
Alex Pardo, Albert Clapes, Sergio Escalera, & Oriol Pujol. (2013). Actions in Context: System for people with Dementia. In 2nd International Workshop on Citizen Sensor Networks (Citisen2013) at the European Conference on Complex Systems (pp. 3–14). Springer International Publishing.
Abstract: In the next forty years, the number of people living with dementia is expected to triple. In the last stages, people affected by this disease become dependent. This hinders the autonomy of the patient and has a huge social impact in time, money and effort. Given this scenario, we propose an ubiquitous system capable of recognizing daily specific actions. The system fuses and synchronizes data obtained from two complementary modalities – ambient and egocentric. The ambient approach consists in a fixed RGB-Depth camera for user and object recognition and user-object interaction, whereas the egocentric point of view is given by a personal area network (PAN) formed by a few wearable sensors and a smartphone, used for gesture recognition. The system processes multi-modal data in real-time, performing paralleled task recognition and modality synchronization, showing high performance recognizing subjects, objects, and interactions, showing its reliability to be applied in real case scenarios.
Keywords: Multi-modal data Fusion; Computer vision; Wearable sensors; Gesture recognition; Dementia
|
|
|
|
David Pujol Perich, Albert Clapes, & Sergio Escalera. (2023). SADA: Semantic adversarial unsupervised domain adaptation for Temporal Action Localization.
Abstract: Temporal Action Localization (TAL) is a complex task that poses relevant challenges, particularly when attempting to generalize on new -- unseen -- domains in real-world applications. These scenarios, despite realistic, are often neglected in the literature, exposing these solutions to important performance degradation. In this work, we tackle this issue by introducing, for the first time, an approach for Unsupervised Domain Adaptation (UDA) in sparse TAL, which we refer to as Semantic Adversarial unsupervised Domain Adaptation (SADA). Our contributions are threefold: (1) we pioneer the development of a domain adaptation model that operates on realistic sparse action detection benchmarks; (2) we tackle the limitations of global-distribution alignment techniques by introducing a novel adversarial loss that is sensitive to local class distributions, ensuring finer-grained adaptation; and (3) we present a novel set of benchmarks based on EpicKitchens100 and CharadesEgo, that evaluate multiple domain shifts in a comprehensive manner. Our experiments indicate that SADA improves the adaptation across domains when compared to fully supervised state-of-the-art and alternative UDA methods, attaining a performance boost of up to 6.14% mAP.
|
|
|
|
Victor Ponce, Baiyu Chen, Marc Oliu, Ciprian Corneanu, Albert Clapes, Isabelle Guyon, et al. (2016). ChaLearn LAP 2016: First Round Challenge on First Impressions – Dataset and Results. In 14th European Conference on Computer Vision Workshops.
Abstract: This paper summarizes the ChaLearn Looking at People 2016 First Impressions challenge data and results obtained by the teams in the rst round of the competition. The goal of the competition was to automatically evaluate ve \apparent“ personality traits (the so-called \Big Five”) from videos of subjects speaking in front of a camera, by using human judgment. In this edition of the ChaLearn challenge, a novel data set consisting of 10,000 shorts clips from YouTube videos has been made publicly available. The ground truth for personality traits was obtained from workers of Amazon Mechanical Turk (AMT). To alleviate calibration problems between workers, we used pairwise comparisons between videos, and variable levels were reconstructed by tting a Bradley-Terry-Luce model with maximum likelihood. The CodaLab open source
platform was used for submission of predictions and scoring. The competition attracted, over a period of 2 months, 84 participants who are grouped in several teams. Nine teams entered the nal phase. Despite the diculty of the task, the teams made great advances in this round of the challenge.
Keywords: Behavior Analysis; Personality Traits; First Impressions
|
|
|
|
Simeon Petkov, Xavier Carrillo, Petia Radeva, & Carlo Gatta. (2014). Diaphragm border detection in coronary X-ray angiographies: New method and applications. CMIG - Computerized Medical Imaging and Graphics, 38(4), 296–305.
Abstract: X-ray angiography is widely used in cardiac disease diagnosis during or prior to intravascular interventions. The diaphragm motion and the heart beating induce gray-level changes, which are one of the main obstacles in quantitative analysis of myocardial perfusion. In this paper we focus on detecting the diaphragm border in both single images or whole X-ray angiography sequences. We show that the proposed method outperforms state of the art approaches. We extend a previous publicly available data set, adding new ground truth data. We also compose another set of more challenging images, thus having two separate data sets of increasing difficulty. Finally, we show three applications of our method: (1) a strategy to reduce false positives in vessel enhanced images; (2) a digital diaphragm removal algorithm; (3) an improvement in Myocardial Blush Grade semi-automatic estimation.
|
|
|
|
Hugo Prol, Vincent Dumoulin, & Luis Herranz. (2018). Cross-Modulation Networks for Few-Shot Learning.
Abstract: A family of recent successful approaches to few-shot learning relies on learning an embedding space in which predictions are made by computing similarities between examples. This corresponds to combining information between support and query examples at a very late stage of the prediction pipeline. Inspired by this observation, we hypothesize that there may be benefits to combining the information at various levels of abstraction along the pipeline. We present an architecture called Cross-Modulation Networks which allows support and query examples to interact throughout the feature extraction process via a feature-wise modulation mechanism. We adapt the Matching Networks architecture to take advantage of these interactions and show encouraging initial results on miniImageNet in the 5-way, 1-shot setting, where we close the gap with state-of-the-art.
|
|
|
|
Olivier Penacchio, Laura Dempere-Marco, & Xavier Otazu. (2012). Switching off brightness induction through induction-reversed images. In Perception (Vol. 41, 208).
Abstract: Brightness induction is the modulation of the perceived intensity of an
area by the luminance of surrounding areas. Although V1 is traditionally regarded as
an area mostly responsive to retinal information, neurophysiological evidence
suggests that it may explicitly represent brightness information. In this work, we
investigate possible neural mechanisms underlying brightness induction. To this end,
we consider the model by Z Li (1999 Computation and Neural Systems10187-212)
which is constrained by neurophysiological data and focuses on the part of V1
responsible for contextual influences. This model, which has proven to account for
phenomena such as contour detection and preattentive segmentation, shares with
brightness induction the relevant effect of contextual influences. Importantly, the
input to our network model derives from a complete multiscale and multiorientation
wavelet decomposition, which makes it possible to recover an image reflecting the
perceived luminance and successfully accounts for well known psychophysical
effects for both static and dynamic contexts. By further considering inverse problem
techniques we define induction-reversed images: given a target image, we build an
image whose perceived luminance matches the actual luminance of the original
stimulus, thus effectively canceling out brightness induction effects. We suggest that
induction-reversed images may help remove undesired perceptual effects and can
find potential applications in fields such as radiological image interpretation
|
|
|
|
Olivier Penacchio, Laura Dempere-Marco, & Xavier Otazu. (2012). A Neurodynamical Model Of Brightness Induction In V1 Following Static And Dynamic Contextual Influences. In 8th Federation of European Neurosciences (Vol. 6, pp. 63–64).
Abstract: Brightness induction is the modulation of the perceived intensity of an area by the luminance of surrounding areas. Although striate cortex is traditionally regarded as an area mostly responsive to ensory (i.e. retinal) information,
neurophysiological evidence suggests that perceived brightness information mightbe explicitly represented in V1.
Such evidence has been observed both in anesthetised cats where neuronal response modulations have been found to follow luminance changes outside the receptive felds and in human fMRI measurements. In this work, possible neural mechanisms that ofer a plausible explanation for such phenomenon are investigated. To this end, we consider the model proposed by Z.Li (Li, Network:Comput. Neural Syst., 10 (1999)) which is based on neurophysiological evidence and focuses on the part of V1 responsible for contextual infuences, i.e. layer 2-3 pyramidal cells, interneurons, and horizontal intracortical connections. This model has reproduced other phenomena such as contour detection and preattentive segmentation, which share with brightness induction the relevant efect of contextual infuences. We have extended the original model such that the input to the network is obtained from a complete multiscale and multiorientation wavelet decomposition, thereby allowing the recovery of an image refecting the perceived intensity. The proposed model successfully accounts for well known psychophysical efects for static contexts (among them: the White's and modifed White's efects, the Todorovic, Chevreul, achromatic ring patterns, and grating induction efects) and also for brigthness induction in dynamic contexts defned by modulating the luminance of surrounding areas (e.g. the brightness of a static central area is perceived to vary in antiphase to the sinusoidal luminance changes of its surroundings). This work thus suggests that intra-cortical interactions in V1 could partially explain perceptual brightness induction efects and reveals how a common general architecture may account for several different fundamental processes emerging early in the visual processing pathway.
|
|
|
|
Giuseppe Pezzano, Oliver Diaz, Vicent Ribas Ripoll, & Petia Radeva. (2021). CoLe-CNN+: Context learning – Convolutional neural network for COVID-19-Ground-Glass-Opacities detection and segmentation. CBM - Computers in Biology and Medicine, 136, 104689.
Abstract: The most common tool for population-wide COVID-19 identification is the Reverse Transcription-Polymerase Chain Reaction test that detects the presence of the virus in the throat (or sputum) in swab samples. This test has a sensitivity between 59% and 71%. However, this test does not provide precise information regarding the extension of the pulmonary infection. Moreover, it has been proven that through the reading of a computed tomography (CT) scan, a clinician can provide a more complete perspective of the severity of the disease. Therefore, we propose a comprehensive system for fully-automated COVID-19 detection and lesion segmentation from CT scans, powered by deep learning strategies to support decision-making process for the diagnosis of COVID-19.
|
|
|
|
Mateusz Pyla, Kamil Deja, Bartłomiej Twardowski, & Tomasz Trzcinski. (2023). Bayesian Flow Networks in Continual Learning.
Abstract: Bayesian Flow Networks (BFNs) has been recently proposed as one of the most promising direction to universal generative modelling, having ability to learn any of the data type. Their power comes from the expressiveness of neural networks and Bayesian inference which make them suitable in the context of continual learning. We delve into the mechanics behind BFNs and conduct the experiments to empirically verify the generative capabilities on non-stationary data.
|
|
|
|
Xavier Perez Sala, Sergio Escalera, Cecilio Angulo, & Jordi Gonzalez. (2014). A survey on model based approaches for 2D and 3D visual human pose recovery. SENS - Sensors, 14(3), 4189–4210.
Abstract: Human Pose Recovery has been studied in the field of Computer Vision for the last 40 years. Several approaches have been reported, and significant improvements have been obtained in both data representation and model design. However, the problem of Human Pose Recovery in uncontrolled environments is far from being solved. In this paper, we define a general taxonomy to group model based approaches for Human Pose Recovery, which is composed of five main modules: appearance, viewpoint, spatial relations, temporal consistence, and behavior. Subsequently, a methodological comparison is performed following the proposed taxonomy, evaluating current SoA approaches in the aforementioned five group categories. As a result of this comparison, we discuss the main advantages and drawbacks of the reviewed literature.
Keywords: human pose recovery; human body modelling; behavior analysis; computer vision
|
|
|
|
Victor Ponce, Sergio Escalera, & Xavier Baro. (2013). Multi-modal Social Signal Analysis for Predicting Agreement in Conversation Settings. In 15th ACM International Conference on Multimodal Interaction (pp. 495–502).
Abstract: In this paper we present a non-invasive ambient intelligence framework for the analysis of non-verbal communication applied to conversational settings. In particular, we apply feature extraction techniques to multi-modal audio-RGB-depth data. We compute a set of behavioral indicators that define communicative cues coming from the fields of psychology and observational methodology. We test our methodology over data captured in victim-offender mediation scenarios. Using different state-of-the-art classification approaches, our system achieve upon 75% of recognition predicting agreement among the parts involved in the conversations, using as ground truth the experts opinions.
|
|
|
|
Cristina Palmero, Jordi Esquirol, Vanessa Bayo, Miquel Angel Cos, Pouya Ahmadmonfared, Joan Salabert, et al. (2017). Automatic Sleep System Recommendation by Multi-modal RBG-Depth-Pressure Anthropometric Analysis. IJCV - International Journal of Computer Vision, 122(2), 212–227.
Abstract: This paper presents a novel system for automatic sleep system recommendation using RGB, depth and pressure information. It consists of a validated clinical knowledge-based model that, along with a set of prescription variables extracted automatically, obtains a personalized bed design recommendation. The automatic process starts by performing multi-part human body RGB-D segmentation combining GrabCut, 3D Shape Context descriptor and Thin Plate Splines, to then extract a set of anthropometric landmark points by applying orthogonal plates to the segmented human body. The extracted variables are introduced to the computerized clinical model to calculate body circumferences, weight, morphotype and Body Mass Index categorization. Furthermore, pressure image analysis is performed to extract pressure values and at-risk points, which are also introduced to the model to eventually obtain the final prescription of mattress, topper, and pillow. We validate the complete system in a set of 200 subjects, showing accurate category classification and high correlation results with respect to manual measures.
Keywords: Sleep system recommendation; RGB-Depth data Pressure imaging; Anthropometric landmark extraction; Multi-part human body segmentation
|
|
