Archives for 2017

Today I read a paper titled “Online Action Recognition based on Incremental Learning of Weighted Covariance Descriptors”

The abstract is:
Online action recognition aims to recognize actions from unsegmented streams of data in a continuous manner.

One of the challenges in online recognition is the accumulation of evidence for decision making.

This paper presents a fast and efficient online method to recognize actions from a stream of noisy skeleton data.

The method adopts a covariance descriptor calculated from skeleton data and is based on a novel method developed for incrementally learning the covariance descriptors, referred to as weighted covariance descriptors, so that past frames have less contributions to the descriptor and current frames and informative frames such as key frames contributes more towards the descriptor.

The online recognition is achieved using an efficient nearest neighbour search against a set of trained actions.

Experimental results on MSRC-12 Kinect Gesture dataset and our newly collocated online action recognition dataset have demonstrated the efficacy of the proposed method.

Today I read a paper titled “Generating Natural Questions About an Image”

The abstract is:
There has been an explosion of work in the vision & language community during the past few years from image captioning to video transcription, and answering questions about images.

These tasks have focused on literal descriptions of the image.

To move beyond the literal, we choose to explore how questions about an image are often directed at commonsense inference and the abstract events evoked by objects in the image.

In this paper, we introduce the novel task of Visual Question Generation (VQG), where the system is tasked with asking a natural and engaging question when shown an image.

We provide three datasets which cover a variety of images from object-centric to event-centric, with considerably more abstract training data than provided to state-of-the-art captioning systems thus far.

We train and test several generative and retrieval models to tackle the task of VQG.

Evaluation results show that while such models ask reasonable questions for a variety of images, there is still a wide gap with human performance which motivates further work on connecting images with commonsense knowledge and pragmatics.

Our proposed task offers a new challenge to the community which we hope furthers interest in exploring deeper connections between vision & language.

Today I read a paper titled “Building Machines That Learn and Think Like People”

The abstract is:
Recent progress in artificial intelligence (AI) has renewed interest in building systems that learn and think like people.

Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats humans in some respects.

Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways.

We review progress in cognitive science suggesting that truly human-like learning and thinking machines will have to reach beyond current engineering trends in both what they learn, and how they learn it.

Specifically, we argue that these machines should (a) build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; (b) ground learning in intuitive theories of physics and psychology, to support and enrich the knowledge that is learned; and (c) harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations.

We suggest concrete challenges and promising routes towards these goals that can combine the strengths of recent neural network advances with more structured cognitive models.

Today I read a paper titled “Combat Models for RTS Games”

The abstract is:
Game tree search algorithms, such as Monte Carlo Tree Search (MCTS), require access to a forward model (or “simulator”) of the game at hand.

However, in some games such forward model is not readily available.

This paper presents three forward models for two-player attrition games, which we call “combat models”, and show how they can be used to simulate combat in RTS games.

We also show how these combat models can be learned from replay data.

We use StarCraft as our application domain.

We report experiments comparing our combat models predicting a combat output and their impact when used for tactical decisions during a real game.

Today I read a paper titled “Memcomputing and Swarm Intelligence”

The abstract is:
We explore the relation between memcomputing, namely computing with and in memory, and swarm intelligence algorithms.

In particular, we show that one can design memristive networks to solve short-path optimization problems that can also be solved by ant-colony algorithms.

By employing appropriate memristive elements one can demonstrate an almost one-to-one correspondence between memcomputing and ant colony optimization approaches.

However, the memristive network has the capability of finding the solution in one deterministic step, compared to the stochastic multi-step ant colony optimization.

This result paves the way for nanoscale hardware implementations of several swarm intelligence algorithms that are presently explored, from scheduling problems to robotics.

Today I read a paper titled “Watch-Bot: Unsupervised Learning for Reminding Humans of Forgotten Actions”

The abstract is:
We present a robotic system that watches a human using a Kinect v2 RGB-D sensor, detects what he forgot to do while performing an activity, and if necessary reminds the person using a laser pointer to point out the related object.

Our simple setup can be easily deployed on any assistive robot.

Our approach is based on a learning algorithm trained in a purely unsupervised setting, which does not require any human annotations.

This makes our approach scalable and applicable to variant scenarios.

Our model learns the action/object co-occurrence and action temporal relations in the activity, and uses the learned rich relationships to infer the forgotten action and the related object.

We show that our approach not only improves the unsupervised action segmentation and action cluster assignment performance, but also effectively detects the forgotten actions on a challenging human activity RGB-D video dataset.

In robotic experiments, we show that our robot is able to remind people of forgotten actions successfully.

Today I read a paper titled “3D Texture Coordinates on Polygon Mesh Sequences”

The abstract is:
A method for creating 3D texture coordinates for a sequence of polygon meshes with changing topology and vertex motion vectors.

Today I read a paper titled “Invariant EKF Design for Scan Matching-aided Localization”

The abstract is:
Localization in indoor environments is a technique which estimates the robot’s pose by fusing data from onboard motion sensors with readings of the environment, in our case obtained by scan matching point clouds captured by a low-cost Kinect depth camera.

We develop both an Invariant Extended Kalman Filter (IEKF)-based and a Multiplicative Extended Kalman Filter (MEKF)-based solution to this problem.

The two designs are successfully validated in experiments and demonstrate the advantage of the IEKF design.

Today I read a paper titled “Low-Cost Scene Modeling using a Density Function Improves Segmentation Performance”

The abstract is:
We propose a low cost and effective way to combine a free simulation software and free CAD models for modeling human-object interaction in order to improve human & object segmentation.

It is intended for research scenarios related to safe human-robot collaboration (SHRC) and interaction (SHRI) in the industrial domain.

The task of human and object modeling has been used for detecting activity, and for inferring and predicting actions, different from those works, we do human and object modeling in order to learn interactions in RGB-D data for improving segmentation.

For this purpose, we define a novel density function to model a three dimensional (3D) scene in a virtual environment (VREP).

This density function takes into account various possible configurations of human-object and object-object relationships and interactions governed by their affordances.

Using this function, we synthesize a large, realistic and highly varied synthetic RGB-D dataset that we use for training.

We train a random forest classifier, and the pixelwise predictions obtained is integrated as a unary term in a pairwise conditional random fields (CRF).

Our evaluation shows that modeling these interactions improves segmentation performance by ~7\% in mean average precision and recall over state-of-the-art methods that ignore these interactions in real-world data.

Our approach is computationally efficient, robust and can run real-time on consumer hardware.

Today I read a paper titled “A Dynamic Boundary Guarding Problem with Translating Targets”

The abstract is:
We introduce a problem in which a service vehicle seeks to guard a deadline (boundary) from dynamically arriving mobile targets.

The environment is a rectangle and the deadline is one of its edges.

Targets arrive continuously over time on the edge opposite the deadline, and move towards the deadline at a fixed speed.

The goal for the vehicle is to maximize the fraction of targets that are captured before reaching the deadline.

We consider two cases; when the service vehicle is faster than the targets, and; when the service vehicle is slower than the targets.

In the first case we develop a novel vehicle policy based on computing longest paths in a directed acyclic graph.

We give a lower bound on the capture fraction of the policy and show that the policy is optimal when the distance between the target arrival edge and deadline becomes very large.

We present numerical results which suggest near optimal performance away from this limiting regime.

In the second case, when the targets are slower than the vehicle, we propose a policy based on servicing fractions of the translational minimum Hamiltonian path.

In the limit of low target speed and high arrival rate, the capture fraction of this policy is within a small constant factor of the optimal.

Today I read a paper titled “Automatic Identification of Animal Breeds and Species Using Bioacoustics and Artificial Neural Networks”

The abstract is:
In this research endeavor, it was hypothesized that the sound produced by animals during their vocalizations can be used as identifiers of the animal breed or species even if they sound the same to unaided human ear.

To test this hypothesis, three artificial neural networks (ANNs) were developed using bioacoustics properties as inputs for the respective automatic identification of 13 bird species, eight dog breeds, and 11 frog species.

Recorded vocalizations of these animals were collected and processed using several known signal processing techniques to convert the respective sounds into computable bioacoustics values.

The converted values of the vocalizations, together with the breed or species identifications, were used to train the ANNs following a ten-fold cross validation technique.

Tests show that the respective ANNs can correctly identify 71.43\% of the birds, 94.44\% of the dogs, and 90.91\% of the frogs.

This result show that bioacoustics and ANN can be used to automatically determine animal breeds and species, which together could be a promising automated tool for animal identification, biodiversity determination, animal conservation, and other animal welfare efforts.

Today I read a paper titled “The eel-like robot”

The abstract is:
The aim of this project is to design, study and build an “eel-like robot” prototype able to swim in three dimensions.

The study is based on the analysis of eel swimming and results in the realization of a prototype with 12 vertebrae, a skin and a head with two fins.

To reach these objectives, a multidisciplinary group of teams and laboratories has been formed in the framework of two French projects.

Today I read a paper titled “Hand Segmentation for Hand-Object Interaction from Depth map”

The abstract is:
Hand-object interaction is important for many applications such as augmented reality, medical application, and human-robot interaction.

Hand segmentation is a necessary pre-process to estimate hand pose and to recognize hand gesture or object in interaction.

However, current hand segmentation method for hand-object interaction is based on color information which is not robust to objects with skin color, skin pigment difference, and light condition variations.

Therefore, we propose the first hand segmentation method for hand-object interaction using depth map.

This is challenging because of only small depth difference between hand and object during interaction.

The proposed method includes two-stage randomized decision forest (RDF) with validation process, bilateral filtering, decision adjustment, and post-processing.

We demonstrate the effectiveness of the proposed method by testing for five objects.

The proposed method achieves the average F1 score of 0.8826 using different model for each object and 0.8645 using a global model for entire objects.

Also, the method takes only about 10ms to process each frame.

We believe that this is the state-of-the-art hand segmentation algorithm using depth map for hand-object interaction.

Today I read a paper titled “Multi-command Chest Tactile Brain Computer Interface for Small Vehicle Robot Navigation”

The abstract is:
The presented study explores the extent to which tactile stimuli delivered to five chest positions of a healthy user can serve as a platform for a brain computer interface (BCI) that could be used in an interactive application such as robotic vehicle operation.

The five chest locations are used to evoke tactile brain potential responses, thus defining a tactile brain computer interface (tBCI).

Experimental results with five subjects performing online tBCI provide a validation of the chest location tBCI paradigm, while the feasibility of the concept is illuminated through information-transfer rates.

Additionally an offline classification improvement with a linear SVM classifier is presented through the case study.