Archives for 2016

Today I read a paper titled “Synchronization and Collision Avoidance in Non-Linear Flocking Networks of Autonomous Agents”

The abstract is:
We introduce and discuss two novel second-order consensus networks with state-dependent couplings of Cucker-Smale type.

The first scheme models flocking to synchronization over a network of agents where the alignment of the agent’s states occurs over a non-trivial limit orbit that is generated by the internal dynamics of each individual agent.

The second scheme models the speed alignment of a group of agents which avoid approaching each other closer than a prescribed distance.

While seemingly different, both of these systems can be analyzed using the same mathematical methods.

We rigorously analyze both examples and reveal their striking similarities.

We arrive at sufficient conditions that relate the initial configurations and the systems’ parameters that give rise to a collective common behavior.

Simulation examples are presented to support our theoretical conclusions.

Today I read a paper titled “3-D Hand Pose Estimation from Kinect’s Point Cloud Using Appearance Matching”

The abstract is:
We present a novel appearance-based approach for pose estimation of a human hand using the point clouds provided by the low-cost Microsoft Kinect sensor.

Both the free-hand case, in which the hand is isolated from the surrounding environment, and the hand-object case, in which the different types of interactions are classified, have been considered.

The hand-object case is clearly the most challenging task having to deal with multiple tracks.

The approach proposed here belongs to the class of partial pose estimation where the estimated pose in a frame is used for the initialization of the next one.

The pose estimation is obtained by applying a modified version of the Iterative Closest Point (ICP) algorithm to synthetic models to obtain the rigid transformation that aligns each model with respect to the input data.

The proposed framework uses a “pure” point cloud as provided by the Kinect sensor without any other information such as RGB values or normal vector components.

For this reason, the proposed method can also be applied to data obtained from other types of depth sensor, or RGB-D camera.

Today I read a paper titled “Virtual World, Defined from a Technological Perspective, and Applied to Video Games, Mixed Reality and the Metaverse”

The abstract is:
There is no generally accepted definition for a virtual world, with many complimentary terms and acronyms having emerged implying a virtual world.

Advances in systems architecture techniques such as, host migration of instances, mobile ad-hoc networking, and distributed computing, bring in to question whether those architectures can actually support a virtual world.

Without a concrete definition, controversy ensues and it is problematic to design an architecture for a virtual world.

Several researchers provided a definition but aspects of each definition are still problematic and simply can not be applied to contemporary technologies.

The approach of this article is to sample technologies using grounded theory, and obtain a definition for a `virtual world’ that is directly applicable to technology.

The obtained definition is compared with related work and used to classify advanced technologies, such as: a pseudo-persistent video game, a MANet, virtual and mixed reality, and the Metaverse.

The results of this article include: a break down of which properties set apart the various technologies; a definition that is validated by comparing it with other definitions; an ontology showing the relation of the different complimentary terms and acronyms; and, the usage of pseudo-persistence to categories those technologies which only mimic persistence.

Today I read a paper titled “Reconfiguration of 3D Crystalline Robots Using O(log n) Parallel Moves”

The abstract is:
We consider the theoretical model of Crystalline robots, which have been introduced and prototyped by the robotics community.

These robots consist of independently manipulable unit-square atoms that can extend/contract arms on each side and attach/detach from neighbors.

These operations suffice to reconfigure between any two given (connected) shapes.

The worst-case number of sequential moves required to transform one connected configuration to another is known to be Theta(n).

However, in principle, atoms can all move simultaneously.

We develop a parallel algorithm for reconfiguration that runs in only O(log n) parallel steps, although the total number of operations increases slightly to Theta(nlogn).

The result is the first (theoretically) almost-instantaneous universally reconfigurable robot built from simple units.

Today I read a paper titled “Hands-free Evolution of 3D-printable Objects via Eye Tracking”

The abstract is:
Interactive evolution has shown the potential to create amazing and complex forms in both 2-D and 3-D settings.

However, the algorithm is slow and users quickly become fatigued.

We propose that the use of eye tracking for interactive evolution systems will both reduce user fatigue and improve evolutionary success.

We describe a systematic method for testing the hypothesis that eye tracking driven interactive evolution will be a more successful and easier-to-use design method than traditional interactive evolution methods driven by mouse clicks.

We provide preliminary results that support the possibility of this proposal, and lay out future work to investigate these advantages in extensive clinical trials.

Today I read a paper titled “Stitching Stabilizer: Two-frame-stitching Video Stabilization for Embedded Systems”

The abstract is:
In conventional electronic video stabilization, the stabilized frame is obtained by cropping the input frame to cancel camera shake.

While a small cropping size results in strong stabilization, it does not provide us satisfactory results from the viewpoint of image quality, because it narrows the angle of view.

By fusing several frames, we can effectively expand the area of input frames, and achieve strong stabilization even with a large cropping size.

Several methods for doing so have been studied.

However, their computational costs are too high for embedded systems such as smartphones.

We propose a simple, yet surprisingly effective algorithm, called the stitching stabilizer.

It stitches only two frames together with a minimal computational cost.

It can achieve real-time processes in embedded systems, for Full HD and 30 FPS videos.

To clearly show the effect, we apply it to hyperlapse.

Using several clips, we show it produces more strongly stabilized and natural results than the existing solutions from Microsoft and Instagram.

Today I read a paper titled “Text Matching as Image Recognition”

The abstract is:
Matching two texts is a fundamental problem in many natural language processing tasks.

An effective way is to extract meaningful matching patterns from words, phrases, and sentences to produce the matching score.

Inspired by the success of convolutional neural network in image recognition, where neurons can capture many complicated patterns based on the extracted elementary visual patterns such as oriented edges and corners, we propose to model text matching as the problem of image recognition.

Firstly, a matching matrix whose entries represent the similarities between words is constructed and viewed as an image.

Then a convolutional neural network is utilized to capture rich matching patterns in a layer-by-layer way.

We show that by resembling the compositional hierarchies of patterns in image recognition, our model can successfully identify salient signals such as n-gram and n-term matchings.

Experimental results demonstrate its superiority against the baselines.

Today I read a paper titled “Simulation of Color Blindness and a Proposal for Using Google Glass as Color-correcting Tool”

The abstract is:
The human visual color response is driven by specialized cells called cones, which exist in three types, viz.

R, G, and B.

Software is developed to simulate how color images are displayed for different types of color blindness.

Specified the default color deficiency associated with a user, it generates a preview of the rainbow (in the visible range, from red to violet) and shows up, side by side with a colorful image provided as input, the display correspondent colorblind.

The idea is to provide an image processing after image acquisition to enable a better perception ofcolors by the color blind.

Examples of pseudo-correction are shown for the case of Protanopia (red blindness).

The system is adapted into a screen of an i-pad or a cellphone in which the colorblind observe the camera, the image processed with color detail previously imperceptible by his naked eye.

As prospecting, wearable computer glasses could be manufactured to provide a corrected image playback.

The approach can also provide augmented reality for human vision by adding the UV or IR responses as a new feature of Google Glass.

Today I read a paper titled “Neural Aggregation Network for Video Face Recognition”

The abstract is:
In this paper, we present a Neural Aggregation Network (NAN) for video face recognition.

The network takes a face video or face image set of a person with variable number of face frames as its input, and produces a compact and fixed-dimension visual representation of that person.

The whole network is composed of two modules.

The feature embedding module is a CNN which maps each face frame into a feature representation.

The neural aggregation module is composed of two content based attention blocks which is driven by a memory storing all the features extracted from the face video through the feature embedding module.

The output of the first attention block adapts the second, whose output is adopted as the aggregated representation of the video faces.

Due to the attention mechanism, this representation is invariant to the order of the face frames.

The experiments show that the proposed NAN consistently outperforms hand-crafted aggregations such as average pooling, and achieves state-of-the-art accuracy on three video face recognition datasets: the YouTube Face, IJB-A and Celebrity-1000 datasets.

Today I read a paper titled “Function-Described Graphs for Structural Pattern Recognition”

The abstract is:
We present in this article the model Function-described graph (FDG), which is a type of compact representation of a set of attributed graphs (AGs) that borrow from Random Graphs the capability of probabilistic modelling of structural and attribute information.

We define the FDGs, their features and two distance measures between AGs (unclassified patterns) and FDGs (models or classes) and we also explain an efficient matching algorithm.

Two applications of FDGs are presented: in the former, FDGs are used for modelling and matching 3D-objects described by multiple views, whereas in the latter, they are used for representing and recognising human faces, described also by several views.

Today I read a paper titled “Multicommand Tactile Brain Computer Interface based on Fingertips or Head Stimulation”

The abstract is:
The paper presents results from a computational neuroscience study conducted to test vibrotactile stimuli delivered to subject fingertips and head areas in order to evoke the somatosensory brain responses utilized in a haptic brain computer interface (hBCI) paradigm.

We present the preliminary and very encouraging results, with subjects conducting online hBCI interfacing experiments, ranging from 40% to 90% with a very fast inter-stimulus-interval (ISI) of 250ms.

The presented results confirm our hypothesis that the hBCI paradigm concept is valid and it allows for rapid stimuli presentation in order to achieve a satisfactory information-transfer-rate of the novel BCI.

Today I read a paper titled “Affordable Virtual Reality System Architecture for Representation of Implicit Object Properties”

The abstract is:
A flexible, scalable and affordable virtual reality software system architecture is proposed.

This solution can be easily implemented on different hardware configurations: on a single computer or on a computer cluster.

The architecture is aimed to be integrated in the workflow for solving engineering tasks and oriented towards presenting implicit object properties through multiple sensorial channels (visual, audio and haptic).

Implicit properties represent hidden object features (i.e.

magnetization, radiation, humidity, toxicity, etc.) which cannot be perceived by the observer through his or her senses but require specialized equipment in order to expand the sensory ability of the observer.

Our approach extends the underlying general scene graph structure incorporating additional effects nodes for implicit properties representation.

Today I read a paper titled “It’s about time: Online Macrotask Sequencing in Expert Crowdsourcing”

The abstract is:
We introduce the problem of Task Assignment and Sequencing (TAS), which adds the timeline perspective to expert crowdsourcing optimization.

Expert crowdsourcing involves macrotasks, like document writing, product design, or web development, which take more time than typical binary microtasks, require expert skills, assume varying degrees of knowledge over a topic, and require crowd workers to build on each other’s contributions.

Current works usually assume offline optimization models, which consider worker and task arrivals known and do not take into account the element of time.

Realistically however, time is critical: tasks have deadlines, expert workers are available only at specific time slots, and worker/task arrivals are not known a-priori.

Our work is the first to address the problem of optimal task sequencing for online, heterogeneous, time-constrained macrotasks.

We propose tas-online, an online algorithm that aims to complete as many tasks as possible within budget, required quality and a given timeline, without future input information regarding job release dates or worker availabilities.

Results, comparing tas-online to four typical benchmarks, show that it achieves more completed jobs, lower flow times and higher job quality.

This work has practical implications for improving the Quality of Service of current crowdsourcing platforms, allowing them to offer cost, quality and time improvements for expert tasks.

Today I read a paper titled “Single-Image Depth Perception in the Wild”

The abstract is:
This paper studies single-image depth perception in the wild, i.e., recovering depth from a single image taken in unconstrained settings.

We introduce a new dataset “Depth in the Wild” consisting of images in the wild annotated with relative depth between pairs of random points.

We also propose a new algorithm that learns to estimate metric depth using annotations of relative depth.

Compared to the state of the art, our algorithm is simpler and performs better.

Experiments show that our algorithm, combined with existing RGB-D data and our new relative depth annotations, significantly improves single-image depth perception in the wild.

Today I read a paper titled “Footprint-Driven Locomotion Composition”

The abstract is:
One of the most efficient ways of generating goal-directed walking motions is synthesising the final motion based on footprints.

Nevertheless, current implementations have not examined the generation of continuous motion based on footprints, where different behaviours can be generated automatically.

Therefore, in this paper a flexible approach for footprint-driven locomotion composition is presented.

The presented solution is based on the ability to generate footprint-driven locomotion, with flexible features like jumping, running, and stair stepping.

In addition, the presented system examines the ability of generating the desired motion of the character based on predefined footprint patterns that determine which behaviour should be performed.

Finally, it is examined the generation of transition patterns based on the velocity of the root and the number of footsteps required to achieve the target behaviour smoothly and naturally.

Today I read a paper titled “Multilinear Wavelets: A Statistical Shape Space for Human Faces”

The abstract is:
We present a statistical model for $3$D human faces in varying expression, which decomposes the surface of the face using a wavelet transform, and learns many localized, decorrelated multilinear models on the resulting coefficients.

Using this model we are able to reconstruct faces from noisy and occluded $3$D face scans, and facial motion sequences.

Accurate reconstruction of face shape is important for applications such as tele-presence and gaming.

The localized and multi-scale nature of our model allows for recovery of fine-scale detail while retaining robustness to severe noise and occlusion, and is computationally efficient and scalable.

We validate these properties experimentally on challenging data in the form of static scans and motion sequences.

We show that in comparison to a global multilinear model, our model better preserves fine detail and is computationally faster, while in comparison to a localized PCA model, our model better handles variation in expression, is faster, and allows us to fix identity parameters for a given subject.

Today I read a paper titled “Motion Planning Of an Autonomous Mobile Robot Using Artificial Neural Network”

The abstract is:
The paper presents the electronic design and motion planning of a robot based on decision making regarding its straight motion and precise turn using Artificial Neural Network (ANN).

The ANN helps in learning of robot so that it performs motion autonomously.

The weights calculated are implemented in microcontroller.

The performance has been tested to be excellent.

Today I read a paper titled “Real-time High Resolution Fusion of Depth Maps on GPU”

The abstract is:
A system for live high quality surface reconstruction using a single moving depth camera on a commodity hardware is presented.

High accuracy and real-time frame rate is achieved by utilizing graphics hardware computing capabilities via OpenCL and by using sparse data structure for volumetric surface representation.

Depth sensor pose is estimated by combining serial texture registration algorithm with iterative closest points algorithm (ICP) aligning obtained depth map to the estimated scene model.

Aligned surface is then fused into the scene.

Kalman filter is used to improve fusion quality.

Truncated signed distance function (TSDF) stored as block-based sparse buffer is used to represent surface.

Use of sparse data structure greatly increases accuracy of scanned surfaces and maximum scanning area.

Traditional GPU implementation of volumetric rendering and fusion algorithms were modified to exploit sparsity to achieve desired performance.

Incorporation of texture registration for sensor pose estimation and Kalman filter for measurement integration improved accuracy and robustness of scanning process.

Today I read a paper titled “Learning to Track at 100 FPS with Deep Regression Networks”

The abstract is:
Machine learning techniques are often used in computer vision due to their ability to leverage large amounts of training data to improve performance.

Unfortunately, most generic object trackers are still trained from scratch online and do not benefit from the large number of videos that are readily available for offline training.

We propose a method for using neural networks to track generic objects in a way that allows them to improve performance by training on labeled videos.

Previous attempts to use neural networks for tracking are very slow to run and not practical for real-time applications.

In contrast, our tracker uses a simple feed-forward network with no online training required, allowing our tracker to run at 100 fps during test time.

Our tracker trains from both labeled video as well as a large collection of images, which helps prevent overfitting.

The tracker learns generic object motion and can be used to track novel objects that do not appear in the training set.

We test our network on a standard tracking benchmark to demonstrate our tracker’s state-of-the-art performance.

Our network learns to track generic objects in real-time as they move throughout the world.

Today I read a paper titled “Detecting Engagement in Egocentric Video”

The abstract is:
In a wearable camera video, we see what the camera wearer sees.

While this makes it easy to know roughly what he chose to look at, it does not immediately reveal when he was engaged with the environment.

Specifically, at what moments did his focus linger, as he paused to gather more information about something he saw? Knowing this answer would benefit various applications in video summarization and augmented reality, yet prior work focuses solely on the “what” question (estimating saliency, gaze) without considering the “when” (engagement).

We propose a learning-based approach that uses long-term egomotion cues to detect engagement, specifically in browsing scenarios where one frequently takes in new visual information (e.g., shopping, touring).

We introduce a large, richly annotated dataset for ego-engagement that is the first of its kind.

Our approach outperforms a wide array of existing methods.

We show engagement can be detected well independent of both scene appearance and the camera wearer’s identity.