We present an approach for real-time, robust and accurate hand pose estimation from moving egocentric RGB-D cameras in cluttered real environments. Existing methods typically fail for hand-object interactions in cluttered scenes imaged from egocentric viewpoints, common for virtual or augmented reality applications. Our approach uses two subsequently applied Convolutional Neural Networks (CNNs) to localize the hand and regress 3D joint locations. Hand localization is achieved by using a CNN to estimate the 2D position of the hand center in the input, even in the presence of clutter and occlusions. The localized hand position, together with the corresponding input depth value, is used to generate a normalized cropped image that is fed into a second CNN to regress relative 3D hand joint locations in real time. For added accuracy, robustness and temporal stability, we refine the pose estimates using a kinematic pose tracking energy. To train the CNNs, we introduce a new photorealistic dataset that uses a merged reality approach to capture and synthesize large amounts of annotated data of natural hand interaction in cluttered scenes. Through quantitative and qualitative evaluation, we show that our method is robust to self-occlusion and occlusions by objects, particularly in moving egocentric perspectives.
In this position paper we discuss three main shortcomings of existing approaches to counterfactual causality from the computer science perspective, and sketch lines of work to try and overcome these issues: (1) causality definitions should be driven by a set of precisely specified requirements rather than specific examples; (2) causality frameworks should support system dynamics; (3) causality analysis should have a well-understood behavior in presence of abstraction.
Mobile robots, be it autonomous or teleoperated, require stable communication with the base station to exchange valuable information. Given the stochastic elements in radio signal propagation, such as shadowing and fading, and the possibilities of unpredictable events or hardware failures, communication loss often presents a significant mission risk, both in terms of probability and impact, especially in Urban Search and Rescue (USAR) operations. Depending on the circumstances, disconnected robots are either abandoned or attempt to autonomously back-trace their way to the base station. Although recent results in Communication-Aware Motion Planning can be used to effectively manage connectivity with robots, there are no results focusing on autonomously re-establishing the wireless connectivity of a mobile robot without back-tracking or using detailed a priori information of the network. In this paper, we present a robust and online radio signal mapping method using Gaussian Random Fields and propose a Resilient Communication-Aware Motion Planner (RCAMP) that integrates the above signal mapping framework with a motion planner. RCAMP considers both the environment and the physical constraints of the robot, based on the available sensory information. We also propose a self-repair strategy using RCMAP, that takes both connectivity and the goal position into account when driving to a connection-safe position in the event of a communication loss. We demonstrate the proposed planner in a set of realistic simulations of an exploration task in single or multi-channel communication scenarios.
Artificial General Intelligence (AGI) or Strong AI aims to create machines with human-like or human-level intelligence, which is still a very ambitious goal when compared to the existing computing and AI systems. After many hype cycles and lessons from AI history, it is clear that a big conceptual leap is needed for crossing the starting line to kick-start mainstream AGI research. This position paper aims to make a small conceptual contribution toward reaching that starting line. After a broad analysis of the AGI problem from different perspectives, a system-theoretic and engineering-based research approach is introduced, which builds upon the existing mainstream AI and systems foundations. Several promising cross-fertilization opportunities between systems disciplines and AI research are identified. Specific potential research directions are discussed.
We present Autonomous Rssi based RElative poSitioning and Tracking (ARREST), a new robotic sensing system for tracking and following a moving, RF-emitting object, which we refer to as the Leader, solely based on signal strength information. This kind of system can expand the horizon of autonomous mobile tracking and distributed robotics into many scenarios with limited visibility such as nighttime, dense forests, and cluttered environments. Our proposed tracking agent, which we refer to as the TrackBot, uses a single rotating, off-the-shelf, directional antenna, novel angle and relative speed estimation algorithms, and Kalman filtering to continually estimate the relative position of the Leader with decimeter level accuracy (which is comparable to a state-of-the-art multiple access point based RF-localization system) and the relative speed of the Leader with accuracy on the order of 1 m/s. The TrackBot feeds the relative position and speed estimates into a Linear Quadratic Gaussian (LQG) controller to generate a set of control outputs to control the orientation and the movement of the TrackBot. We perform an extensive set of real world experiments with a full-fledged prototype to demonstrate that the TrackBot is able to stay within 5m of the Leader with: (1) more than $99\%$ probability in line of sight scenarios, and (2) more than $70\%$ probability in no line of sight scenarios, when it moves 1.8X faster than the Leader. For ground truth estimation in real world experiments, we also developed an integrated TDoA based distance and angle estimation system with centimeter level localization accuracy in line of sight scenarios. While providing a first proof of concept, our work opens the door to future research aimed at further improvements of autonomous RF-based tracking.
Deep learning approaches such as convolutional neural nets have consistently outperformed previous methods on challenging tasks such as dense, semantic segmentation. However, the various proposed networks perform differently, with behaviour largely influenced by architectural choices and training settings. This paper explores Ensembles of Multiple Models and Architectures (EMMA) for robust performance through aggregation of predictions from a wide range of methods. The approach reduces the influence of the meta-parameters of individual models and the risk of overfitting the configuration to a particular database. EMMA can be seen as an unbiased, generic deep learning model which is shown to yield excellent performance, winning the first position in the BRATS 2017 competition among 50+ participating teams.
Discovering statistical structure from links is a fundamental problem in the analysis of social networks. Choosing a misspecified model, or equivalently, an incorrect inference algorithm will result in an invalid analysis or even falsely uncover patterns that are in fact artifacts of the model. This work focuses on unifying two of the most widely used link-formation models: the stochastic blockmodel (SBM) and the small world (or latent space) model (SWM). Integrating techniques from kernel learning, spectral graph theory, and nonlinear dimensionality reduction, we develop the first statistically sound polynomial-time algorithm to discover latent patterns in sparse graphs for both models. When the network comes from an SBM, the algorithm outputs a block structure. When it is from an SWM, the algorithm outputs estimates of each node's latent position.
Global Average Pooling (GAP) [4] has been used previously to generate class activation for image classification tasks. The motivation behind SIMILARnet comes from the fact that the convolutional filters possess position information of the essential features and hence, combination of the feature maps could help us locate the class instances in an image. We propose a biologically inspired model that is free of differential connections and doesn't require separate training thereby reducing computation overhead. Our novel architecture generates promising results and unlike existing methods, the model is not sensitive to the input image size, thus promising wider application. Codes for the experiment and illustrations can be found at: https://github.com/brcsomnath/Advanced-GAP .
Tracking with a Pan-Tilt-Zoom (PTZ) camera has been a research topic in computer vision for many years. Compared to tracking with a still camera, the images captured with a PTZ camera are highly dynamic in nature because the camera can perform large motion resulting in quickly changing capture conditions. Furthermore, tracking with a PTZ camera involves camera control to position the camera on the target. For successful tracking and camera control, the tracker must be fast enough, or has to be able to predict accurately the next position of the target. Therefore, standard benchmarks do not allow to assess properly the quality of a tracker for the PTZ scenario. In this work, we use a virtual PTZ framework to evaluate different tracking algorithms and compare their performances. We also extend the framework to add target position prediction for the next frame, accounting for camera motion and processing delays. By doing this, we can assess if predicting can make long-term tracking more robust as it may help slower algorithms for keeping the target in the field of view of the camera. Results confirm that both speed and robustness are required for tracking under the PTZ scenario.
In this paper, we propose a refined scene text detector with a \textit{novel} Feature Enhancement Network (FEN) for Region Proposal and Text Detection Refinement. Retrospectively, both region proposal with \textit{only} $3\times 3$ sliding-window feature and text detection refinement with \textit{single scale} high level feature are insufficient, especially for smaller scene text. Therefore, we design a new FEN network with \textit{task-specific}, \textit{low} and \textit{high} level semantic features fusion to improve the performance of text detection. Besides, since \textit{unitary} position-sensitive RoI pooling in general object detection is unreasonable for variable text regions, an \textit{adaptively weighted} position-sensitive RoI pooling layer is devised for further enhancing the detecting accuracy. To tackle the \textit{sample-imbalance} problem during the refinement stage, we also propose an effective \textit{positives mining} strategy for efficiently training our network. Experiments on ICDAR 2011 and 2013 robust text detection benchmarks demonstrate that our method can achieve state-of-the-art results, outperforming all reported methods in terms of F-measure.
Transparency, user trust, and human comprehension are popular ethical motivations for interpretable machine learning. In support of these goals, researchers evaluate model explanation performance using humans and real world applications. This alone presents a challenge in many areas of artificial intelligence. In this position paper, we propose a distinction between descriptive and persuasive explanations. We discuss reasoning suggesting that functional interpretability may be correlated with cognitive function and user preferences. If this is indeed the case, evaluation and optimization using functional metrics could perpetuate implicit cognitive bias in explanations that threaten transparency. Finally, we propose two potential research directions to disambiguate cognitive function and explanation models, retaining control over the tradeoff between accuracy and interpretability.
We propose an approach to reduce both computational complexity and data storage requirements for the online positioning stage of a fingerprinting-based indoor positioning system (FIPS) by introducing segmentation of the region of interest (RoI) into sub-regions, sub-region selection using a modified Jaccard index, and feature selection based on randomized least absolute shrinkage and selection operator (LASSO). We implement these steps into a Bayesian framework of position estimation using the maximum a posteriori (MAP) principle. An additional benefit of these steps is that the time for estimating the position, and the required data storage are virtually independent of the size of the RoI and of the total number of available features within the RoI. Thus the proposed steps facilitate application of FIPS to large areas. Results of an experimental analysis using real data collected in an office building using a Nexus 6P smart phone as user device and a total station for providing position ground truth corroborate the expected performance of the proposed approach. The positioning accuracy obtained by only processing 10 automatically identified features instead of all available ones and limiting position estimation to 10 automatically identified sub-regions instead of the entire RoI is equivalent to processing all available data. In the chosen example, 50% of the errors are less than 1.8 m and 90% are less than 5 m. However, the computation time using the automatically identified subset of data is only about 1% of that required for processing the entire data set.
Models that can execute natural language instructions for situated robotic tasks such as assembly and navigation have several useful applications in homes, offices, and remote scenarios. We study the semantics of spatially-referred configuration and arrangement instructions, based on the challenging Bisk-2016 blank-labeled block dataset. This task involves finding a source block and moving it to the target position (mentioned via a reference block and offset), where the blocks have no names or colors and are just referred to via spatial location features. We present novel models for the subtasks of source block classification and target position regression, based on joint-loss language and spatial-world representation learning, as well as CNN-based and dual attention models to compute the alignment between the world blocks and the instruction phrases. For target position prediction, we compare two inference approaches: annealed sampling via policy gradient versus expectation inference via supervised regression. Our models achieve the new state-of-the-art on this task, with an improvement of 47% on source block accuracy and 22% on target position distance.
We propose the cascade attribute learning network (CALNet), which can learn attributes in a control task separately and assemble them together. Our contribution is twofold: first we propose attribute learning in reinforcement learning (RL). Attributes used to be modeled using constraint functions or terms in the objective function, making it hard to transfer. Attribute learning, on the other hand, models these task properties as modules in the policy network. We also propose using novel cascading compensative networks in the CALNet to learn and assemble attributes. Using the CALNet, one can zero shoot an unseen task by separately learning all its attributes, and assembling the attribute modules. We have validated the capacity of our model on a wide variety of control problems with attributes in time, position, velocity and acceleration phases.
Examining locomotion has improved our basic understanding of motor control and aided in treating motor impairment. Mice and rats are premier models of human disease and increasingly the model systems of choice for basic neuroscience. High frame rates (250 Hz) are needed to quantify the kinematics of these running rodents. Manual tracking, especially for multiple markers, becomes time-consuming and impossible for large sample sizes. Therefore, the need for automatic segmentation of these markers has grown in recent years. Here, we address this need by presenting a method to segment the markers using the SLIC superpixel method. The 2D coordinates on the image plane are projected to a 3D domain using direct linear transform (DLT) and a 3D Kalman filter has been used to predict the position of markers based on the speed and position of markers from the previous frames. Finally, a probabilistic function is used to find the best match among superpixels. The method is evaluated for different difficulties for tracking of the markers and it achieves 95% correct labeling of markers.
We present an end-to-end learning method for chess, relying on deep neural networks. Without any a priori knowledge, in particular without any knowledge regarding the rules of chess, a deep neural network is trained using a combination of unsupervised pretraining and supervised training. The unsupervised training extracts high level features from a given position, and the supervised training learns to compare two chess positions and select the more favorable one. The training relies entirely on datasets of several million chess games, and no further domain specific knowledge is incorporated. The experiments show that the resulting neural network (referred to as DeepChess) is on a par with state-of-the-art chess playing programs, which have been developed through many years of manual feature selection and tuning. DeepChess is the first end-to-end machine learning-based method that results in a grandmaster-level chess playing performance.
We present a novel 3D face reconstruction technique that leverages sparse photometric stereo (PS) and latest advances on face registration/modeling from a single image. We observe that 3D morphable faces approach provides a reasonable geometry proxy for light position calibration. Specifically, we develop a robust optimization technique that can calibrate per-pixel lighting direction and illumination at a very high precision without assuming uniform surface albedos. Next, we apply semantic segmentation on input images and the geometry proxy to refine hairy vs. bare skin regions using tailored filters. Experiments on synthetic and real data show that by using a very small set of images, our technique is able to reconstruct fine geometric details such as wrinkles, eyebrows, whelks, pores, etc, comparable to and sometimes surpassing movie quality productions.
We present R-FCN-3000, a large-scale real-time object detector in which objectness detection and classification are decoupled. To obtain the detection score for an RoI, we multiply the objectness score with the fine-grained classification score. Our approach is a modification of the R-FCN architecture in which position-sensitive filters are shared across different object classes for performing localization. For fine-grained classification, these position-sensitive filters are not needed. R-FCN-3000 obtains an mAP of 34.9% on the ImageNet detection dataset and outperforms YOLO-9000 by 18% while processing 30 images per second. We also show that the objectness learned by R-FCN-3000 generalizes to novel classes and the performance increases with the number of training object classes - supporting the hypothesis that it is possible to learn a universal objectness detector. Code will be made available.
Autonomous underwater vehicles (AUVs) have been deployed for underwater exploration. However, its potential is confined by its limited on-board battery energy and data storage capacity. This problem has been addressed using docking systems by underwater recharging and data transfer for AUVs. In this work, we propose a vision based framework for underwater docking following these systems. The proposed framework comprises two modules; (i) a detection module which provides location information on underwater docking stations in 2D images captured by an on-board camera, and (ii) a pose estimation module which recovers the relative 3D position and orientation between docking stations and AUVs from the 2D images. For robust and credible detection of docking stations, we propose a convolutional neural network called Docking Neural Network (DoNN). For accurate pose estimation, a perspective-n-point algorithm is integrated into our framework. In order to examine our framework in underwater docking tasks, we collected a dataset of 2D images, named Underwater Docking Images Dataset (UDID), in an experimental water pool. To the best of our knowledge, UDID is the first publicly available underwater docking dataset. In the experiments, we first evaluate performance of the proposed detection module on UDID and its deformed variations. Next, we assess the accuracy of the pose estimation module by ground experiments, since it is not feasible to obtain true relative position and orientation between docking stations and AUVs under water. Then, we examine the pose estimation module by underwater experiments in our experimental water pool. Experimental results show that the proposed framework can be used to detect docking stations and estimate their relative pose efficiently and successfully, compared to the state-of-the-art baseline systems.
Image splicing detection is of fundamental importance in digital forensics and therefore has attracted increasing attention recently. In this paper, a color image splicing detection approach is proposed based on Markov transition probability of quaternion component separation in quaternion discrete cosine transform (QDCT) domain and quaternion wavelet transform (QWT) domain. Firstly, Markov features of the intra-block and inter-block between block QDCT coefficients are obtained from the real part and three imaginary parts of QDCT coefficients respectively. Then, additional Markov features are extracted from luminance (Y) channel in quaternion wavelet transform domain to characterize the dependency of position among quaternion wavelet subband coefficients. Finally, ensemble classifier (EC) is exploited to classify the spliced and authentic color images. The experiment results demonstrate that the proposed approach can outperforms some state-of-the-art methods.