Transfer Learning for Human & AI

Optimal $k$-thresholding algorithms overcome the shortcomings of traditional hard thresholding algorithms . Theory presents that the transition point of thenumber of measurements is on the order of $k \log({en}/{k)$ The algorithms can achieve linear convergence when the number of measurements required for successful recovery has a negative correlation with number of iterations and the algorithms achieve linearconvergence .…

We develop an algorithm that computes strongly continuous semigroups on infinite-dimensional Hilbert spaces with explicit error control . The algorithm is based on a combination of a regularizedfunctional calculus, suitable contour quadrature rules, and adaptivecomputation of resolvents in infinite dimensions .…

Transfer learning avoids training a new model from scratch by transferring knowledge from a source task to a target task . As many source models are available, it is difficult for Datascientists to select the best source model for the target task manually .…

Long-term object detection requires the integration of frame-based resultsover several seconds . For non-deformable objects, long-term detection is oftenaddressed using object detection followed by video tracking . Unfortunately,tracking is inapplicable to objects that undergo dramatic changes in appearance from frame to frame .…

The main process behind a software defect is that an error-prone scenariotriggers human error modes, which psychologists have observed to recur acrossdiverse activities . The proposed ideaemphasizes predicting the exact location and form of a possible defect . The approach was able to predict, atthe requirement phase, the exact locations and forms of 7 out of the 22 (31.8%) specific types of defects that were found in the code .…

Neural Radiance Fields (NeRF) have recently demonstrated photo-realistic results for the task of novel view synthesis . In this paper, we propose to apply novel view synthesive view synthesis to the robot relocalization problem . We demonstrateimprovement of camera pose regression thanks to an additional synthetic dataset .…

Adaptive SGD is an adaptive elasticmodel averaging stochastic gradient descent algorithm for heterogeneousmulti-GPUs . It is characterized by dynamic scheduling, adaptive batch sizescaling, and normalized model merging . It outperforms four state-of-the-art solutions in time-to-accuracy and is scalable with the number of GPUs .…

Using DeepVecFont, for the first time, vector glyphs whose quality and compactness are bothcomparable to human-designed ones can be automatically generated . The key idea is to adopt the techniques of image synthesis, sequencemodeling and differentiable rasterization to exhaustively exploit thedual-modality information (i.e.,…

We study the convergence of a symmetrized Graph Laplacian matrix induced by a Gaussian kernel evaluated on pairs of embedded data, sampled from a manifold with boundary . We deduce theconvergence of the \emph{truncated Graph LaPlacian}, which is recentlynumerically observed in applications, and provide a detailed numerical investigation on simple manifolds .…

A high-order gas-kinetic scheme is developed for the equation of radiation hydrodynamics in equilibrium-diffusion limit . An implicit-explicit scheme is applied in which the hydrodynamic part is treated explicitly and the radiationdiffusion is treated implicitly . For the radiation diffusionterm, the nonlinear generalized minimal residual (GMRES) method is used .…

A generative model can be used to build abetter recognizer through the control of content and style . By training our controllable handwriting synthesizer on the same data, we cansynthesize handwriting with previously underrepresented content . Overall, we achieve a 66% reduction in CharacterError Rate.…

In this paper, we consider a directed hypergraph as cooperative network, anddefine the Myerson value for directedhypergraphs . We prove the axiomatization of the value, namely strong component efficiency and fairness . We modified the concept of safety defined by Li-Shan, and proved the condition about the safety of the hyperedge .…

Alibaba serves an indispensable role forhundreds of different travel scenarios from Fliggy, Taobao, Alipay apps, etc. To provide personalized recommendation service for users visiting differentscenarios, there are two critical issues to be carefully addressed . In this paper, we propose a novelScenario-Aware Ranking Network (SAR-Net) to address these issues .…

Hyperspectral imaging has been increasingly used for underwater survey applications . We propose to combine techniques from simultaneous localization and mapping, structure-from-motion and 3Dreconstruction and use them to create 3D models with hyperspectral texture . We show that the proposed method creates highly accurate 3D reconstructions of underwater environments .…

We propose a novel conflict resolution framework for IoT services in multi-resident smart homes . The proposed framework employs a preferenceextraction model based on a temporal proximity strategy . We design a preferenceaggregation model using a matrix factorization-based approach . The concepts of current resident item matrix and idealresident item matrix are introduced as key criteria to cater to the conflictresolution framework .…

Machine Learning (ML) techniques for the next generation wireless networks have shown promising results in the recent years . ML techniqueshave been used for load balancing in Self-Organizing Networks (SON) In the context of load balancing and ML, several studies propose network managementautomation (NMA) from the perspective of a single and centralized agent .…

The approach localizes the fault through symbolic reasoning, and synthesizes apatch such that the repaired program can pass all unit tests . It applies domain-specific abstraction to simplify network data structures and utilizesmodular analysis to facilitate function summary reuse for symbolic analysis .…

We consider the problem of using expert data with unobserved confounders forimitation and reinforcement learning . We propose an adjustment of standard imitation learning algorithms to fit this setup . We also discuss distribution shift between expert data and the online environment when the data is only partially observable .…

Growth in technology has resulted in the large-scale collection and processing of Personally Identifiable Information by organizations that rundigital services such as websites . Several African countries have recently started enacting new dataprotection regulations due to recent technological innovations . Only 16 percent of third-party tracking companies that track users on a particular website are disclosed in the site’s privacy policy statements .…

This paper presents several strategies to tune the parameters of metaheuristics for (discrete) design optimization of reinforced concrete(RC) structures . A novel utility metric is proposed, based on the area under the average performance curve . The simplest strategy is suitable to tune good `generalist’ methods, i.e.,…

We describe a Question Answering (QA) dataset that contains complex questions with conditional answers . ConditionalQA is challenging for many of the existing QA models, especially in selecting answer conditions . We believe that this dataset willmotivate further research in answering complex questions over long documents .…

We analyse an energy minimisation problem recently proposed for modellingsmectic-A liquid crystals . The optimality conditions give a coupled nonlinearsystem of partial differential equations . Our two main results are aproof of the existence of solutions to the minimisation problems and a priori error estimates for its discretisation using the $\mathcal{C}^0$ interior penalty method .…

3D object detection often involves complicated training and testingpipelines, which require substantial domain knowledge about individualdatasets . Inspired by recent non-maximum suppression-free 2D objects detection models, we propose a new architecture on point clouds . This approach aligns the outputs of the teachermodel and the student model in a permutation-invariant fashion, significantly simplifying knowledge distillation for the 3D detection task .…

Conversational Recommendation System (CRS) uses theinteractive form of the dialogue systems to solve the intrinsic problems of traditional recommendation systems . The existing CRS models are unable to deal with theexploitation and exploration (E&E) problem well, resulting in the heavy burdenon users .…

In this paper, we propose a new deep unfolding neural network based on theADMM algorithm for analysis Compressed Sensing . The proposed network jointlylearns a redundant analysis operator for sparsification and reconstructs thesignal of interest . We compare our proposed network with the state-of-the-art state of theart deep unfolding networks, that also learns an orthogonal sparsifier .…

Rendering stable hard surfaces is an important problem in haptics for many tasks, including training simulators for orthopedic surgery or dentistry . Current impedance devices cannot provide enough force and stiffness to render awall . We propose to address these limitations by combininghaptic augmented reality, untethered haptic interaction, and anencountered-type haptic display .…

Simultaneous Localization and Mapping (SLAM) enables autonomous robots tonavigate and execute their tasks through unknown environments . However,performing SLAM in large environments with a single robot is not efficient, and visual or LiDAR-based SLAM requires feature extraction and matching algorithms .…

This paper concerns solving the steady radiative transfer equation using the physics informed neural networks (PINNs) The ideaof PINNs is to minimize a least-square loss function, that consists of theresidual from the governing equation, the mismatch from the boundaryconditions, and other physical constraints such as conservation .…

We present and review an algorithmic and theoretical framework for improving neural network architecture design via momentum . We consider how momentum can improve the architecture design for recurrent neural networks, neural ordinary differential equations (ODEs), and transformers . Weshow that integrating momentum into neural network architectures has several remarkable theoretical and empirical benefits, including overcoming the vanishing gradient issues in training RNNs and neural ODEs, resulting in effective learning long-termdependencies .…

Studies have shown improvement in users’ mental model of the VQA system when they are exposed to examples of how these systems answer certain Image-Question (IQ) pairs . We use recent advances in generativeadversarial networks (GANs) to generate counterfactual images by deleting andinpainting certain regions of interest in the image .…

New concepts from computer science and engineering are often hard to grasp . The Who-wants-to-be-a-millionaire?-style quiz game lets the playerexperience indoor positioning based on Bluetooth signal strength firsthand . We found that such an interactive game demonstrator can function as aconversation-opener and is useful in helping introduce concepts relevant for future jobs .…

Humans have mental models that allow them to plan, experiment, and reason inthe physical world . How should an intelligent agent go about learning such models? In this paper, we will study if models of the world learned in anopen-ended physics environment, without any specific tasks, can be reused for downstream physics reasoning tasks .…

The class of row monomial matrices is closed under multiplication, but not under ordinary matrix addition . The most significant difference is the summation operation . The algebra plays an important role in the study of DFA,especially for synchronizing automata .…

Recently, pretrained generativesequence-to-sequence (seq2seq) models have achieved great success in questionanswering . We propose a novel method for applying Transformer models to extractivequestion answering tasks . Viewing cross-attention as an architectural prior, we apply joint training to further improve QA performance .…

Serverless computing along with Function-as-a-Service (FaaS) are forming a computing paradigm that is anticipated to found the next generation of cloud systems . Serverless cloud systems offer a highly transparent infrastructure that enables user applications to scale in the granularity of their functions .…

Magnetic polarizability tensor (MPT) offers an economical characterisation of metallic objects . MPT spectral signature can be determined frommeasurements of the induced voltage over a range frequencies in a metalsignature for a hidden object . With classification in mind, it can also becomomomputed in advance for different threat and non-threat objects .…

Social engineering scams (SES) has been existed since the adoption of thetelecommunications by humankind . An earlier version of the scams includeleveraging premium phone service to charge the consumers and service providers . There are variety of techniques being considered to scam the people due to the advancements in digital data access capabilities andInternet technology .…

Electroencephalogram (EEG) can objectively reflect emotional state andchanges . But transmission mechanism of EEG in the brain and its relationship with emotion are still ambiguous to human beings . New method, named EEG emotion Transformer (EeT), adapts theconventional Transformer architecture to EEG signals by enabling spatiospectralfeature learning directly from sequences of EEG signals .…

We propose a novel conflict resolution framework for IoT services in multi-resident smart homes . The proposed framework employs a preferenceextraction model based on a temporal proximity strategy . We design a preferenceaggregation model using a matrix factorization-based approach . The concepts of current resident item matrix and idealresident item matrix are introduced as key criteria to cater to the conflictresolution framework .…

Empirical science of neural scaling laws is a rapidly growing area of significant importance to the future of machine learning . Our findings shed new light on therelationship between scale and generalization . Our key observationsare that such performance improvements are well-approximated by power laws(linear log-log plots) as the training set size increases .…

There is a growing interest in estimating the user agreement by considering the standard deviation of the scores, instead of only predicting the mean aestheticopinion score . With such loss, the model is encouraged to learn theuncertainty of the content that is relevant to the diversity of observers’opinions, i.e.,…

Autonomous mobile manipulation robots that can collect trolleys are widely used to liberate human resources and fight epidemics . The proposed system integrates a compact hardwaredesign and a progressive perception and planning framework . For the perception, we first develop a 3D trolley detectionmethod that combines object detection and keypoint estimation .…

The main process behind a software defect is that an error-prone scenariotriggers human error modes, which psychologists have observed to recur acrossdiverse activities . The proposed ideaemphasizes predicting the exact location and form of a possible defect . The approach was able to predict, atthe requirement phase, the exact locations and forms of 7 out of the 22 (31.8%) specific types of defects that were found in the code .…

In this paper, we propose THOMAS, a joint multi-agent trajectory predictionframework . We present a unified model architecture for fast andsimultaneous agent future heatmap estimation leveraging hierarchical and sparseimage generation . We demonstrate that heatmap output enables a higher level ofcontrol on the predicted trajectories compared to vanilla multi-modaltrajectory regression .…

Few-shot class incremental learning is a key challenge for machine learning systems deployed innon-stationary environments . The key to this approach is a new family of subspace regularizationschemes that encourage weight vectors for new classes to lie close to thesubspace spanned by the weights of existing classes .…

Recommendation systems play an important role in today’s digital world . Less research effort has been devoted to physical exercise recommendation systems . Sedentary lifestyles have become the major driver of several diseases as well as healthcare costs . In this paper, we develop a recommendation system for daily exercise activitiesto users based on their history, profile and similar users .…

Linear temporal logic (LTL) is a specification language for finite sequences called traces . It is widely used in program verification, motion planning inrobotics, process mining, and many other areas . Existing solutions suffer from two limitations: they do notscale beyond small formulas, and they may exhaust computational resources .…

Gaussian distribution is widely used in mechanism design for differentialprivacy (DP) Thanks to its sub-Gaussian tail, it significantly reduces the chance of outliers when responding to queries . In practice, this may limit the use of the Gaussian mechanism for large datasets with strong privacy requirements .…

Outfit recommendation often requires a complex and large model that incurs huge memory and time costs . The explosive number of outfit candidates amplifies the data sparsity problem, often leading to poor outfit representation . False NegativeDistillation (FND) exploits false-negative information from the teacher model while not requiring the ranking of all candidates .…

The essence of meaning conveyed by words is best described by a compass-like power-danger framework . Analysis of a disparate collection of large-scale English language corpora –literature, news, Wikipedia, talk radio, and social media — shows that naturallanguage exhibits a systematic bias toward safe, low danger words .…