Deep Deformation Detail Synthesis for Thin Shell Models

In physics-based cloth animation, rich folds and detailed wrinkles areachieved at the cost of expensive computational resources and huge labortuning . However, existing works usually utilize coordinate-based representations which cannot cope with large-scale deformation . We develop a temporally and spatiallyas-consistent-as-possible deformation representation (named TS-ACAP) and aDeformTransformer network to learn the mapping from low-resolution meshes todetailed ones .…

Text to Audio Grounding Building Correspondence Between Captions and Sound Events

Automated Audio Captioning is a cross-modal task, generating natural languagedescriptions to summarize the audio clips’ sound events . This paper contributes an AudioGrounding dataset, which provides correspondence between sound events and captions provided inAudiocaps . A baseline approach is provided, resulting in an event-F1 score of 28.3% and a Polyphonic Sound Detection Score (PSDS) score of 14.7% .…

Structural Similarity of Boundary Conditions and an Efficient Local Search Algorithm for Goal Conflict Identification

In goal-oriented requirements engineering, goal conflict identification is offundamental importance for requirements analysis . The task aims to find the situations which make the goals diverge within the domain, calledboundary conditions (BCs) The existing approaches for goal conflictidentification fail to find sufficient BCs and general BCs which cover morecombinations of circumstances .…

An Interaction aware Evaluation Method for Highly Automated Vehicles

It is important to build a rigorous verification and validation (V&V) processto evaluate the safety of highly automated vehicles before their widedeployment on public roads . In this paper, we propose an interaction-aware framework for HAV safety evaluation which is suitable for somehighly-interactive driving scenarios including highway merging, roundaboutentering, etc.…

Models we Can Trust Toward a Systematic Discipline of Agent Based Model Interpretation and Validation

We advocate the development of a discipline of interacting with and extracting information from models, both mathematical (e.g. game-theoreticones) and computational models . We outline some directions for the development . of a such a discipline: – development of logical frameworks for the systematic formalspecification of stylized facts and social mechanisms in (mathematical andcomputational) social science .…

A Robotic Model of Hippocampal Reverse Replay for Reinforcement Learning

Hippocampal reverse replay is thought to contribute to learning, andparticularly reinforcement learning, in animals . We present a computational model of learning in the hippocampus that builds on a previous model of the hippocampus . We conclude that reverse replay may enhance reinforcement learning inthe mammalian hippocampal-striatal system rather than provide its coremechanism.…

Instance Specific Approximations for Submodular Maximization

For many optimization problems in machine learning, finding an optimalsolution is computationally intractable . The main challenge is that an optimal solution cannot be efficiently computed forintractable problems . A major question is how to measure the performance of an algorithm in comparison to the optimal solution on instances we encounter in practice .…

Experimental Study on Probabilistic ToA and AoA Joint Localization in Real Indoor Environments

Joint localization algorithm significantly outperforms baselines using only ToA or AoA measurements . It achieves 2-D sub-meter accuracy at the90%-ile. We also numerically demonstrate that the algorithmis more robust to synchronization errors than the baseline using ToAmeasurements only. We evaluate the algorithm performance using aproprietary prototype deployed in an indoor factory environment.…

Quantum Entropic Causal Inference

A direct generalization of the existing causal inference techniquesto the quantum domain is not possible due to superposition and entanglement . We unify classical and quantum causal inference in a principled way paving the way for future applications in quantum computing and networking .…

The Power of D hops in Matching Power Law Graphs

Paper studies seeded graph matching for power-law graphs . Algorithm exploits the low-degree seeds in suitably-defined $D$-hop neighborhoods . This significantly reduces the number of seeds needed to trigger a cascading process to match the rest of thegraphs . Our result achieves an exponential reduction inthe seed size requirement, as the best previously known result requires $n^{1/2+\epsilon}$ seeds .…

End to End Dereverberation Beamforming and Speech Recognition with Improved Numerical Stability and Advanced Frontend

Recently, the end-to-end approach has been successfully applied tomulti-speaker speech separation and recognition in both single-channel and multichannel conditions . However, severe performance degradation is stillobserved in the reverberant and noisy scenarios, and there is still a large performance gap between anechoic and reverberant conditions .…

The Curious Case of Integrator Reach Sets Part I Basic Theory

This is the first of a two part paper investigating the geometry of theintegrator reach sets . In this Part I, we establish that this compact convex set is semialgebraic, translated zonoid, andnot a spectrahedron . We also deduce the closedform formula for the volume and diameter of this set, and discuss their scaling with state dimension and time .…

DeepThermal Combustion Optimization for Thermal Power Generating Units Using Offline Reinforcement Learning

Thermal power generation plays a dominant role in the world’s electricity supply . It consumes large amounts of coal worldwide, and causes serious airpollution . At its core, is a new model-based reinforcement learning (RL) framework, called MORE, which leverages historical operational data of a TPGU to solve a highly complex Markov decision process problem via purely offline training .…

Enhanced Modality Transition for Image Captioning

Image captioning model is a cross-modality knowledge discovery task . It aims to automatically describe an image with an informative and coherentsentence . The previous encoder-decoder frameworks forward the visual vectors to the recurrent language model, forcing the recurrent units to generate a sentence based on the visual features .…

Behavioral QLTL

Behavioral QLTL is a variant of linear-time temporal logic on infinite traces with second-order quantifiers . The functions that assign thetruth value of the quantified propositions along the trace can only depend on the past . This gives to the logic a strategic flavor that we usually associate to planning .…

Practical Mutation Testing at Scale

Mutation testing builds on top of mutation analysis and is a testingtechnique that uses mutants as test goals to create or improve a test suite . Google has a codebase of two billion lines of code and more than 500,000,000 tests are executed on adaily basis .…

Progresses and Challenges in Link Prediction

Link prediction is a paradigmatic problem in network science, which aims atestimating the existence likelihoods of nonobserved links, based on knowntopology . This Perspective will summarize representative progresses about local similarity indices, link predictability, network embedding, matrixcompletion, ensemble learning and others .…

Explore the Context Optimal Data Collection for Context Conditional Dynamics Models

In this paper, we learn dynamics models for parametrized families of systems with varying properties . The dynamics models are formulated as stochastic processes conditioned on a latent context variable . The probabilisticformulation allows us to compute an action sequence which, for a limited number of environment interactions, optimally explores the given system within the given family .…

Recurrent Model Predictive Control

Recurrent Model Predictive Control (RMPC) algorithm can converge to the optimal policy by minimizing the designed loss function . The number of prediction steps is equal to the number of recurrent cycles of the learned policy function . We further prove the convergence and optimality of the RMPC algorithm thoroughBellman optimality principle, and demonstrate its generality and efficiencyusing two numerical examples .…

Investigating Local and Global Information for Automated Audio Captioning with Transfer Learning

Automated audio captioning (AAC) aims at generating summarizing descriptions for audio clips . Multitudesinous concepts are described in an audio caption, ranging from local information such as sound events to global information like acoustic scenery . Two source tasks identified to represent local and global information, being Audio Tagging (AT)and Acoustic Scene Classification (ASC) Experiments are conducted on the AACbenchmark dataset Clotho and Audiocaps, amounting to a vast increase in metrics with topic transfer learning .…

Uniform Elgot Iteration in Foundations

Category theory is famous for its innovative way of thinking of concepts by descriptions . We provide ageneric categorical iteration-based notion of partiality . We show that the emerging free structures, which we dubuniform-iteration algebras, enjoy various desirable properties, in particular,yield an equational lifting monad .…

Data to Physicalization A Survey of the Physical Rendering Process

Physical representations of data offer physical and spatial ways of lookingat, navigating, and interacting with data . Rendering in the scope of this research refers to the back-and-forth process from digital design to digital fabrication and its challenges . We developed a corpus of example data physicalizations from research literature and physicalization practice .…

Greedy Multi step Off Policy Reinforcement Learning

Multi-step off-policy reinforcement learning has achieved great success . However, existing multi-step methods usually impose a fixed prior on the bootstrap steps . The new method has two desiredproperties: It can flexibly adjust the bootstrapping step based on the quality of the data and the learned value function .…

An experimental demonstration of the memristor test

A simple and unambiguous test has been suggested [J. Phys. D:Applied Physics, 52, 01LT01 (2018] to check experimentally if a resistor withmemory is indeed a memristor . Such a test would represent the litmus test for claims about memristors (in the ideal sense) But such a test has yet to be applied widely to physical devices .…

Performance Improvement of LoRa Modulation with Signal Combining

Low-power long-range (LoRa) modulation has been used to satisfy the low powerand large coverage requirements of Internet of Things (IoT) networks . In thispaper, we investigate performance improvements of LoRa modulation when agateway is equipped with multiple antennas . We derive the optimal decisionrules for both coherent and non-coherent detections when combining signalsreceived from multiple antennas.…

The SmartSHARK Repository Mining Data

The SmartSHARK repository mining data is a collection of rich and detailed information about the evolution of software projects . The data set provides a rich source of data that enables us to explore research questions that require data from different sources and/or longitudinal data over time .…

Robust k Center with Two Types of Radii

In the non-uniform $k$-center problem, the objective is to cover points in ametric space with specified number of balls of different radii . Chakrabarty,Goyal, and Krishnaswamy [ICALP 2016, Trans. on Algs 2020] (CGK, henceforth)give a constant factor approximation when there are two types of radii.…

Revisiting the Memristor Concept within Basic Circuit Theory

In this paper we revisit the memristor concept within circuit theory . We show that for general memristive systems the $\varphi-q$ curve is not single-valued or not even closed . An approach suitable to explain the inductive behavior of the giant squid axon had already been developed in the 1960s, with the introduction of “time-variant resistors” We also point out the ambiguities resulting from anon rigorous usage of the flux linkage concept .…

Exact epidemic models from a tensor product formulation

A general framework for obtaining exact transition rate matrices forstochastic systems on networks is presented and applied to many well-known models of epidemiology . The state of the population is described as a vector in the tensor product space of $N$ individual probability vectorspaces, whose dimension equals the number of compartments of theepidemiological model $n_c$.…

SeqNet Learning Descriptors for Sequence based Hierarchical Place Recognition

A novel hybrid system creates a high performance initial match hypothesis generator using shortlearnt sequential descriptors . Sequential descriptors are generated using a temporal convolutional network dubbed SeqNet, encoding shortimage sequences using 1-D convolutions . They are then matched against the corresponding temporal descriptors from the reference dataset to provide an ordered list of place match hypotheses .…