|
Marc Bolaños, Alvaro Peris, Francisco Casacuberta, & Petia Radeva. (2017). VIBIKNet: Visual Bidirectional Kernelized Network for Visual Question Answering. In 8th Iberian Conference on Pattern Recognition and Image Analysis.
Abstract: In this paper, we address the problem of visual question answering by proposing a novel model, called VIBIKNet. Our model is based on integrating Kernelized Convolutional Neural Networks and Long-Short Term Memory units to generate an answer given a question about an image. We prove that VIBIKNet is an optimal trade-off between accuracy and computational load, in terms of memory and time consumption. We validate our method on the VQA challenge dataset and compare it to the top performing methods in order to illustrate its performance and speed.
Keywords: Visual Qestion Aswering; Convolutional Neural Networks; Long short-term memory networks
|
|
|
Marc Bolaños, Alvaro Peris, Francisco Casacuberta, Sergi Solera, & Petia Radeva. (2018). Egocentric video description based on temporally-linked sequences. JVCIR - Journal of Visual Communication and Image Representation, 50, 205–216.
Abstract: Egocentric vision consists in acquiring images along the day from a first person point-of-view using wearable cameras. The automatic analysis of this information allows to discover daily patterns for improving the quality of life of the user. A natural topic that arises in egocentric vision is storytelling, that is, how to understand and tell the story relying behind the pictures.
In this paper, we tackle storytelling as an egocentric sequences description problem. We propose a novel methodology that exploits information from temporally neighboring events, matching precisely the nature of egocentric sequences. Furthermore, we present a new method for multimodal data fusion consisting on a multi-input attention recurrent network. We also release the EDUB-SegDesc dataset. This is the first dataset for egocentric image sequences description, consisting of 1,339 events with 3,991 descriptions, from 55 days acquired by 11 people. Finally, we prove that our proposal outperforms classical attentional encoder-decoder methods for video description.
Keywords: egocentric vision; video description; deep learning; multi-modal learning
|
|
|
Marc Bolaños, Maite Garolera, & Petia Radeva. (2014). Video Segmentation of Life-Logging Videos. In 8th Conference on Articulated Motion and Deformable Objects (Vol. 8563, pp. 1–9).
|
|
|
Marc Bolaños, Maite Garolera, & Petia Radeva. (2015). Object Discovery using CNN Features in Egocentric Videos. In Pattern Recognition and Image Analysis, Proceedings of 7th Iberian Conference , ibPRIA 2015 (Vol. 9117, pp. 67–74). LNCS.
Abstract: Lifelogging devices based on photo/video are spreading faster everyday. This growth can represent great benefits to develop methods for extraction of meaningful information about the user wearing the device and his/her environment. In this paper, we propose a semi-supervised strategy for easily discovering objects relevant to the person wearing a first-person camera. The egocentric video sequence acquired by the camera, uses both the appearance extracted by means of a deep convolutional neural network and an object refill methodology that allow to discover objects even in case of small amount of object appearance in the collection of images. We validate our method on a sequence of 1000 egocentric daily images and obtain results with an F-measure of 0.5, 0.17 better than the state of the art approach.
Keywords: Object discovery; Egocentric videos; Lifelogging; CNN
|
|
|
Marc Bolaños, Maite Garolera, & Petia Radeva. (2013). Active labeling application applied to food-related object recognition. In 5th International Workshop on Multimedia for Cooking & Eating Activities (pp. 45–50).
Abstract: Every day, lifelogging devices, available for recording different aspects of our daily life, increase in number, quality and functions, just like the multiple applications that we give to them. Applying wearable devices to analyse the nutritional habits of people is a challenging application based on acquiring and analyzing life records in long periods of time. However, to extract the information of interest related to the eating patterns of people, we need automatic methods to process large amount of life-logging data (e.g. recognition of food-related objects). Creating a rich set of manually labeled samples to train the algorithms is slow, tedious and subjective. To address this problem, we propose a novel method in the framework of Active Labeling for construct- ing a training set of thousands of images. Inspired by the hierarchical sampling method for active learning [6], we propose an Active forest that organizes hierarchically the data for easy and fast labeling. Moreover, introducing a classifier into the hierarchical structures, as well as transforming the feature space for better data clustering, additionally im- prove the algorithm. Our method is successfully tested to label 89.700 food-related objects and achieves significant reduction in expert time labelling.
Active labeling application applied to food-related object recognition ResearchGate. Available from: http://www.researchgate.net/publication/262252017Activelabelingapplicationappliedtofood-relatedobjectrecognition [accessed Jul 14, 2015].
|
|
|
Marc Bolaños, Mariella Dimiccoli, & Petia Radeva. (2017). Towards Storytelling from Visual Lifelogging: An Overview. THMS - IEEE Transactions on Human-Machine Systems, 47(1), 77–90.
Abstract: Visual lifelogging consists of acquiring images that capture the daily experiences of the user by wearing a camera over a long period of time. The pictures taken offer considerable potential for knowledge mining concerning how people live their lives, hence, they open up new opportunities for many potential applications in fields including healthcare, security, leisure and
the quantified self. However, automatically building a story from a huge collection of unstructured egocentric data presents major challenges. This paper provides a thorough review of advances made so far in egocentric data analysis, and in view of the current state of the art, indicates new lines of research to move us towards storytelling from visual lifelogging.
|
|
|
Marc Bolaños, & Petia Radeva. (2016). Simultaneous Food Localization and Recognition. In 23rd International Conference on Pattern Recognition.
Abstract: CoRR abs/1604.07953
The development of automatic nutrition diaries, which would allow to keep track objectively of everything we eat, could enable a whole new world of possibilities for people concerned about their nutrition patterns. With this purpose, in this paper we propose the first method for simultaneous food localization and recognition. Our method is based on two main steps, which consist in, first, produce a food activation map on the input image (i.e. heat map of probabilities) for generating bounding boxes proposals and, second, recognize each of the food types or food-related objects present in each bounding box. We demonstrate that our proposal, compared to the most similar problem nowadays – object localization, is able to obtain high precision and reasonable recall levels with only a few bounding boxes. Furthermore, we show that it is applicable to both conventional and egocentric images.
|
|
|
Marc Bolaños, R. Mestre, Estefania Talavera, Xavier Giro, & Petia Radeva. (2015). Visual Summary of Egocentric Photostreams by Representative Keyframes. In IEEE International Conference on Multimedia and Expo ICMEW2015 (pp. 1–6).
Abstract: Building a visual summary from an egocentric photostream captured by a lifelogging wearable camera is of high interest for different applications (e.g. memory reinforcement). In this paper, we propose a new summarization method based on keyframes selection that uses visual features extracted bymeans of a convolutional neural network. Our method applies an unsupervised clustering for dividing the photostreams into events, and finally extracts the most relevant keyframe for each event. We assess the results by applying a blind-taste test on a group of 20 people who assessed the quality of the
summaries.
Keywords: egocentric; lifelogging; summarization; keyframes
|
|
|
Marc Castello, Jordi Gonzalez, Ariel Amato, Pau Baiget, Carles Fernandez, Josep M. Gonfaus, et al. (2013). Exploiting Multimodal Interaction Techniques for Video-Surveillance. In Multimodal Interaction in Image and Video Applications Intelligent Systems Reference Library (Vol. 48, pp. 135–151). Springer Berlin Heidelberg.
Abstract: In this paper we present an example of a video surveillance application that exploits Multimodal Interactive (MI) technologies. The main objective of the so-called VID-Hum prototype was to develop a cognitive artificial system for both the detection and description of a particular set of human behaviours arising from real-world events. The main procedure of the prototype described in this chapter entails: (i) adaptation, since the system adapts itself to the most common behaviours (qualitative data) inferred from tracking (quantitative data) thus being able to recognize abnormal behaviors; (ii) feedback, since an advanced interface based on Natural Language understanding allows end-users the communicationwith the prototype by means of conceptual sentences; and (iii) multimodality, since a virtual avatar has been designed to describe what is happening in the scene, based on those textual interpretations generated by the prototype. Thus, the MI methodology has provided an adequate framework for all these cooperating processes.
|
|
|
Marc Masana. (2020). Lifelong Learning of Neural Networks: Detecting Novelty and Adapting to New Domains without Forgetting (Joost Van de Weijer, & Andrew Bagdanov, Eds.). Ph.D. thesis, Ediciones Graficas Rey, .
Abstract: Computer vision has gone through considerable changes in the last decade as neural networks have come into common use. As available computational capabilities have grown, neural networks have achieved breakthroughs in many computer vision tasks, and have even surpassed human performance in others. With accuracy being so high, focus has shifted to other issues and challenges. One research direction that saw a notable increase in interest is on lifelong learning systems. Such systems should be capable of efficiently performing tasks, identifying and learning new ones, and should moreover be able to deploy smaller versions of themselves which are experts on specific tasks. In this thesis, we contribute to research on lifelong learning and address the compression and adaptation of networks to small target domains, the incremental learning of networks faced with a variety of tasks, and finally the detection of out-of-distribution samples at inference time.
We explore how knowledge can be transferred from large pretrained models to more task-specific networks capable of running on smaller devices by extracting the most relevant information. Using a pretrained model provides more robust representations and a more stable initialization when learning a smaller task, which leads to higher performance and is known as domain adaptation. However, those models are too large for certain applications that need to be deployed on devices with limited memory and computational capacity. In this thesis we show that, after performing domain adaptation, some learned activations barely contribute to the predictions of the model. Therefore, we propose to apply network compression based on low-rank matrix decomposition using the activation statistics. This results in a significant reduction of the model size and the computational cost.
Like human intelligence, machine intelligence aims to have the ability to learn and remember knowledge. However, when a trained neural network is presented with learning a new task, it ends up forgetting previous ones. This is known as catastrophic forgetting and its avoidance is studied in continual learning. The work presented in this thesis extensively surveys continual learning techniques and presents an approach to avoid catastrophic forgetting in sequential task learning scenarios. Our technique is based on using ternary masks in order to update a network to new tasks, reusing the knowledge of previous ones while not forgetting anything about them. In contrast to earlier work, our masks are applied to the activations of each layer instead of the weights. This considerably reduces the number of parameters to be added for each new task. Furthermore, the analysis on a wide range of work on incremental learning without access to the task-ID, provides insight on current state-of-the-art approaches that focus on avoiding catastrophic forgetting by using regularization, rehearsal of previous tasks from a small memory, or compensating the task-recency bias.
Neural networks trained with a cross-entropy loss force the outputs of the model to tend toward a one-hot encoded vector. This leads to models being too overly confident when presented with images or classes that were not present in the training distribution. The capacity of a system to be aware of the boundaries of the learned tasks and identify anomalies or classes which have not been learned yet is key to lifelong learning and autonomous systems. In this thesis, we present a metric learning approach to out-of-distribution detection that learns the task at hand on an embedding space.
|
|
|
Marc Masana, Bartlomiej Twardowski, & Joost Van de Weijer. (2020). On Class Orderings for Incremental Learning. In ICML Workshop on Continual Learning.
Abstract: The influence of class orderings in the evaluation of incremental learning has received very little attention. In this paper, we investigate the impact of class orderings for incrementally learned classifiers. We propose a method to compute various orderings for a dataset. The orderings are derived by simulated annealing optimization from the confusion matrix and reflect different incremental learning scenarios, including maximally and minimally confusing tasks. We evaluate a wide range of state-of-the-art incremental learning methods on the proposed orderings. Results show that orderings can have a significant impact on performance and the ranking of the methods.
|
|
|
Marc Masana, Idoia Ruiz, Joan Serrat, Joost Van de Weijer, & Antonio Lopez. (2018). Metric Learning for Novelty and Anomaly Detection. In 29th British Machine Vision Conference.
Abstract: When neural networks process images which do not resemble the distribution seen during training, so called out-of-distribution images, they often make wrong predictions, and do so too confidently. The capability to detect out-of-distribution images is therefore crucial for many real-world applications. We divide out-of-distribution detection between novelty detection ---images of classes which are not in the training set but are related to those---, and anomaly detection ---images with classes which are unrelated to the training set. By related we mean they contain the same type of objects, like digits in MNIST and SVHN. Most existing work has focused on anomaly detection, and has addressed this problem considering networks trained with the cross-entropy loss. Differently from them, we propose to use metric learning which does not have the drawback of the softmax layer (inherent to cross-entropy methods), which forces the network to divide its prediction power over the learned classes. We perform extensive experiments and evaluate both novelty and anomaly detection, even in a relevant application such as traffic sign recognition, obtaining comparable or better results than previous works.
|
|
|
Marc Masana, Joost Van de Weijer, & Andrew Bagdanov. (2016). On-the-fly Network pruning for object detection. In International conference on learning representations.
Abstract: Object detection with deep neural networks is often performed by passing a few
thousand candidate bounding boxes through a deep neural network for each image.
These bounding boxes are highly correlated since they originate from the same
image. In this paper we investigate how to exploit feature occurrence at the image scale to prune the neural network which is subsequently applied to all bounding boxes. We show that removing units which have near-zero activation in the image allows us to significantly reduce the number of parameters in the network. Results on the PASCAL 2007 Object Detection Challenge demonstrate that up to 40% of units in some fully-connected layers can be entirely eliminated with little change in the detection result.
|
|
|
Marc Masana, Joost Van de Weijer, Luis Herranz, Andrew Bagdanov, & Jose Manuel Alvarez. (2017). Domain-adaptive deep network compression. In 17th IEEE International Conference on Computer Vision.
Abstract: Deep Neural Networks trained on large datasets can be easily transferred to new domains with far fewer labeled examples by a process called fine-tuning. This has the advantage that representations learned in the large source domain can be exploited on smaller target domains. However, networks designed to be optimal for the source task are often prohibitively large for the target task. In this work we address the compression of networks after domain transfer.
We focus on compression algorithms based on low-rank matrix decomposition. Existing methods base compression solely on learned network weights and ignore the statistics of network activations. We show that domain transfer leads to large shifts in network activations and that it is desirable to take this into account when compressing.
We demonstrate that considering activation statistics when compressing weights leads to a rank-constrained regression problem with a closed-form solution. Because our method takes into account the target domain, it can more optimally
remove the redundancy in the weights. Experiments show that our Domain Adaptive Low Rank (DALR) method significantly outperforms existing low-rank compression techniques. With our approach, the fc6 layer of VGG19 can be compressed more than 4x more than using truncated SVD alone – with only a minor or no loss in accuracy. When applied to domain-transferred networks it allows for compression down to only 5-20% of the original number of parameters with only a minor drop in performance.
|
|
|
Marc Masana, Tinne Tuytelaars, & Joost Van de Weijer. (2021). Ternary Feature Masks: zero-forgetting for task-incremental learning. In 34th IEEE Conference on Computer Vision and Pattern Recognition Workshops (pp. 3565–3574).
Abstract: We propose an approach without any forgetting to continual learning for the task-aware regime, where at inference the task-label is known. By using ternary masks we can upgrade a model to new tasks, reusing knowledge from previous tasks while not forgetting anything about them. Using masks prevents both catastrophic forgetting and backward transfer. We argue -- and show experimentally -- that avoiding the former largely compensates for the lack of the latter, which is rarely observed in practice. In contrast to earlier works, our masks are applied to the features (activations) of each layer instead of the weights. This considerably reduces the number of mask parameters for each new task; with more than three orders of magnitude for most networks. The encoding of the ternary masks into two bits per feature creates very little overhead to the network, avoiding scalability issues. To allow already learned features to adapt to the current task without changing the behavior of these features for previous tasks, we introduce task-specific feature normalization. Extensive experiments on several finegrained datasets and ImageNet show that our method outperforms current state-of-the-art while reducing memory overhead in comparison to weight-based approaches.
|
|