Abstract: State-of-the-art systems on cognitive surveillance identify and describe complex events in selected domains, thus providing end-users with tools to easily access the contents of massive video footage. Nevertheless, as the complexity of events increases in semantics and the types of indoor/outdoor scenarios diversify, it becomes difficult to assess which events describe better the scene, and how to model them at a pixel level to fulfill natural language requests. We present an ontology-based methodology that guides the identification, step-by-step modeling, and generalization of the most relevant events to a specific domain. Our approach considers three steps: (1) end-users provide textual evidence from surveilled video sequences; (2) transcriptions are analyzed top-down to build the knowledge bases for event description; and (3) the obtained models are used to generalize event detection to different image sequences from the surveillance domain. This framework produces user-oriented knowledge that improves on existing advanced interfaces for video indexing and retrieval, by determining the best suited events for video understanding according to end-users. We have conducted experiments with outdoor and indoor scenes showing thefts, chases, and vandalism, demonstrating the feasibility and generalization of this proposal.

Abstract: Having artificial agents to autonomously produce human-like behaviour is one of the most ambitious original goals of Artificial Intelligence (AI) and remains an open problem nowadays. The imitation game originally proposed by Turing constitute a very effective method to prove the indistinguishability of an artificial agent. The behaviour of an agent is said to be indistinguishable from that of a human when observers (the so-called judges in the Turing test) cannot tell apart humans and non-human agents. Different environments, testing protocols, scopes and problem domains can be established to develop limited versions or variants of the original Turing test. In this paper we use a specific version of the Turing test, based on the international BotPrize competition, built in a First-Person Shooter video game, where both human players and non-player characters interact in complex virtual environments. Based on our past experience both in the BotPrize competition and other robotics and computer game AI applications we have developed three new more advanced controllers for believable agents: two based on a combination of the CERA–CRANIUM and SOAR cognitive architectures and other based on ADANN, a system for the automatic evolution and adaptation of artificial neural networks. These two new agents have been put to the test jointly with CCBot3, the winner of BotPrize 2010 competition (Arrabales et al., 2012), and have showed a significant improvement in the humanness ratio. Additionally, we have confronted all these bots to both First-person believability assessment (BotPrize original judging protocol) and Third-person believability assessment, demonstrating that the active involvement of the judge has a great impact in the recognition of human-like behaviour.

Abstract: Article ID 491819
Social Signal Processing is an emergent area of research that focuses on the analysis of social constructs. Dominance and interest are two of these social constructs. Dominance refers to the level of influence a person has in a conversation. Interest, when referred in terms of group interactions, can be defined as the degree of engagement that the members of a group collectively display during their interaction. In this paper, we argue that only using behavioral motion information, we are able to predict the interest of observers when looking at face-to-face interactions as well as the dominant people. First, we propose a simple set of movement-based features from body, face, and mouth activity in order to define a higher set of interaction indicators. The considered indicators are manually annotated by observers. Based on the opinions obtained, we define an automatic binary dominance detection problem and a multiclass interest quantification problem. Error-Correcting Output Codes framework is used to learn to rank the perceived observer's interest in face-to-face interactions meanwhile Adaboost is used to solve the dominant detection problem. The automatic system shows good correlation between the automatic categorization results and the manual ranking made by the observers in both dominance and interest detection problems.

Abstract: We present a novel method for human action recognition (HAR) based on estimated poses from image sequences. We use 3D human pose data as additional information and propose a compact human pose representation, called a weak pose, in a low-dimensional space while still keeping the most discriminative information for a given pose. With predicted poses from image features, we map the problem from image feature space to pose space, where a Bag of Poses (BOP) model is learned for the final goal of HAR. The BOP model is a modified version of the classical bag of words pipeline by building the vocabulary based on the most representative weak poses for a given action. Compared with the standard k-means clustering, our vocabulary selection criteria is proven to be more efficient and robust against the inherent challenges of action recognition. Moreover, since for action recognition the ordering of the poses is discriminative, the BOP model incorporates temporal information: in essence, groups of consecutive poses are considered together when computing the vocabulary and assignment. We tested our method on two well-known datasets: HumanEva and IXMAS, to demonstrate that weak poses aid to improve action recognition accuracies. The proposed method is scene-independent and is comparable with the state-of-art method.

Abstract: The determination of precise skin lesion boundaries in dermoscopic images using automated methods faces many challenges, most importantly, the presence of hair, inconspicuous lesion edges and low contrast in dermoscopic images, and variability in the color, texture and shapes of skin lesions. Existing deep learning-based skin lesion segmentation algorithms are expensive in terms of computational time and memory. Consequently, running such segmentation algorithms requires a powerful GPU and high bandwidth memory, which are not available in dermoscopy devices. Thus, this article aims to achieve precise skin lesion segmentation with minimum resources: a lightweight, efficient generative adversarial network (GAN) model called SLSNet, which combines 1-D kernel factorized networks, position and channel attention, and multiscale aggregation mechanisms with a GAN model. The 1-D kernel factorized network reduces the computational cost of 2D filtering. The position and channel attention modules enhance the discriminative ability between the lesion and non-lesion feature representations in spatial and channel dimensions, respectively. A multiscale block is also used to aggregate the coarse-to-fine features of input skin images and reduce the effect of the artifacts. SLSNet is evaluated on two publicly available datasets: ISBI 2017 and the ISIC 2018. Although SLSNet has only 2.35 million parameters, the experimental results demonstrate that it achieves segmentation results on a par with the state-of-the-art skin lesion segmentation methods with an accuracy of 97.61%, and Dice and Jaccard similarity coefficients of 90.63% and 81.98%, respectively. SLSNet can run at more than 110 frames per second (FPS) in a single GTX1080Ti GPU, which is faster than well-known deep learning-based image segmentation models, such as FCN. Therefore, SLSNet can be used for practical dermoscopic applications.

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.