arXiv subjects

Tianyu Zhang

Publications and source records attributed to Tianyu Zhang.

180 records · Page 10Linked to original sources

arXiv2021

On the Optimality of Nuclear-norm-based Matrix Completion for Problems with Smooth Non-linear Structure

Originally developed for imputing missing entries in low rank, or approximately low rank matrices, matrix completion has proven widely effective in many problems where there is no reason to assume low-dimensional linear structure in the underlying matrix, as would be imposed by rank constraints. In this manuscript, we build some theoretical intuition for this behavior. We consider matrices which are not necessarily low-rank, but lie in a low-dimensional non-linear manifold. We show that nuclear-norm penalization is still effective for recovering these matrices when observations are missing completely at random. In particular, we give upper bounds on the rate of convergence as a function of the number of rows, columns, and observed entries in the matrix, as well as the smoothness and dimension of the non-linear embedding. We additionally give a minimax lower bound: This lower bound agrees with our upper bound (up to a logarithmic factor), which shows that nuclear-norm penalization is (up to log terms) minimax rate optimal for these problems.

math.ST↗

arXiv2021

An Online Projection Estimator for Nonparametric Regression in Reproducing Kernel Hilbert Spaces

The goal of nonparametric regression is to recover an underlying regression function from noisy observations, under the assumption that the regression function belongs to a pre-specified infinite dimensional function space. In the online setting, when the observations come in a stream, it is generally computationally infeasible to refit the whole model repeatedly. There are as of yet no methods that are both computationally efficient and statistically rate-optimal. In this paper, we propose an estimator for online nonparametric regression. Notably, our estimator is an empirical risk minimizer (ERM) in a deterministic linear space, which is quite different from existing methods using random features and functional stochastic gradient. Our theoretical analysis shows that this estimator obtains rate-optimal generalization error when the regression function is known to live in a reproducing kernel Hilbert space. We also show, theoretically and empirically, that the computational expense of our estimator is much lower than other rate-optimal estimators proposed for this online setting.

stat.ME↗

arXiv2021

Spatiotemporal Transformer for Video-based Person Re-identification

Recently, the Transformer module has been transplanted from natural language processing to computer vision. This paper applies the Transformer to video-based person re-identification, where the key issue is to extract the discriminative information from a tracklet. We show that, despite the strong learning ability, the vanilla Transformer suffers from an increased risk of over-fitting, arguably due to a large number of attention parameters and insufficient training data. To solve this problem, we propose a novel pipeline where the model is pre-trained on a set of synthesized video data and then transferred to the downstream domains with the perception-constrained Spatiotemporal Transformer (STT) module and Global Transformer (GT) module. The derived algorithm achieves significant accuracy gain on three popular video-based person re-identification benchmarks, MARS, DukeMTMC-VideoReID, and LS-VID, especially when the training and testing data are from different domains. More importantly, our research sheds light on the application of the Transformer on highly-structured visual data.

cs.CV↗

arXiv2021

Ternary Hashing

This paper proposes a novel ternary hash encoding for learning to hash methods, which provides a principled more efficient coding scheme with performances better than those of the state-of-the-art binary hashing counterparts. Two kinds of axiomatic ternary logic, Kleene logic and Łukasiewicz logic are adopted to calculate the Ternary Hamming Distance (THD) for both the learning/encoding and testing/querying phases. Our work demonstrates that, with an efficient implementation of ternary logic on standard binary machines, the proposed ternary hashing is compared favorably to the binary hashing methods with consistent improvements of retrieval mean average precision (mAP) ranging from 1\% to 5.9\% as shown in CIFAR10, NUS-WIDE and ImageNet100 datasets.

cs.AI↗

arXiv2021

A robust solution of a statistical inverse problem in multiscale computational mechanics using an artificial neural network

This work addresses the inverse identification of apparent elastic properties of random heterogeneous materials using machine learning based on artificial neural networks. The proposed neural network-based identification method requires the construction of a database from which an artificial neural network can be trained to learn the nonlinear relationship between the hyperparameters of a prior stochastic model of the random compliance field and some relevant quantities of interest of an ad hoc multiscale computational model. An initial database made up with input and target data is first generated from the computational model, from which a processed database is deduced by conditioning the input data with respect to the target data using the nonparametric statistics. Two-and three-layer feedforward artificial neural networks are then trained from each of the initial and processed databases to construct an algebraic representation of the nonlinear mapping between the hyperparameters (network outputs) and the quantities of interest (network inputs). The performances of the trained artificial neural networks are analyzed in terms of mean squared error, linear regression fit and probability distribution between network outputs and targets for both databases. An ad hoc probabilistic model of the input random vector is finally proposed in order to take into account uncertainties on the network input and to perform a robustness analysis of the network output with respect to the input uncertainties level. The capability of the proposed neural network-based identification method to efficiently solve the underlying statistical inverse problem is illustrated through two numerical examples developed within the framework of 2D plane stress linear elasticity, namely a first validation example on synthetic data obtained through computational simulations and a second application example on real experimental data obtained through a physical experiment monitored by digital image correlation on a real heterogeneous biological material (beef cortical bone).

cs.LG↗

arXiv2020

UnrealPerson: An Adaptive Pipeline towards Costless Person Re-identification

The main difficulty of person re-identification (ReID) lies in collecting annotated data and transferring the model across different domains. This paper presents UnrealPerson, a novel pipeline that makes full use of unreal image data to decrease the costs in both the training and deployment stages. Its fundamental part is a system that can generate synthesized images of high-quality and from controllable distributions. Instance-level annotation goes with the synthesized data and is almost free. We point out some details in image synthesis that largely impact the data quality. With 3,000 IDs and 120,000 instances, our method achieves a 38.5% rank-1 accuracy when being directly transferred to MSMT17. It almost doubles the former record using synthesized data and even surpasses previous direct transfer records using real data. This offers a good basis for unsupervised domain adaption, where our pre-trained model is easily plugged into the state-of-the-art algorithms towards higher accuracy. In addition, the data distribution can be flexibly adjusted to fit some corner ReID scenarios, which widens the application of our pipeline. We will publish our data synthesis toolkit and synthesized data in https://github.com/FlyHighest/UnrealPerson.

cs.CV↗

arXiv2020

Identifying Grey-box Thermal Models with Bayesian Neural Networks

Smart thermostats are one of the most prevalent home automation products. They learn occupant preferences and schedules, and utilize an accurate thermal model to reduce the energy use of heating and cooling equipment while maintaining the temperature for maximum comfort. Despite the importance of having an accurate thermal model for the operation of smart thermostats, fast and reliable identification of this model is still an open problem. In this paper, we explore various techniques for establishing a suitable thermal model using time series data generated by smart thermostats. We show that Bayesian neural networks can be used to estimate parameters of a grey-box thermal model if sufficient training data is available, and this model outperforms several black-box models in terms of the temperature prediction accuracy. Leveraging real data from 8,884 homes equipped with smart thermostats, we discuss how the prior knowledge about the model parameters can be utilized to quickly build an accurate thermal model for another home with similar floor area and age in the same climate zone. Moreover, we investigate how to adapt the model originally built for the same home in another season using a small amount of data collected in this season. Our results confirm that maintaining only a small number of pre-trained thermal models will suffice to quickly build accurate thermal models for many other homes, and that 1~day smart thermostat data could significantly improve the accuracy of transferred models in another season.

eess.SY↗

arXiv2020

Rethinking the Distribution Gap of Person Re-identification with Camera-based Batch Normalization

The fundamental difficulty in person re-identification (ReID) lies in learning the correspondence among individual cameras. It strongly demands costly inter-camera annotations, yet the trained models are not guaranteed to transfer well to previously unseen cameras. These problems significantly limit the application of ReID. This paper rethinks the working mechanism of conventional ReID approaches and puts forward a new solution. With an effective operator named Camera-based Batch Normalization (CBN), we force the image data of all cameras to fall onto the same subspace, so that the distribution gap between any camera pair is largely shrunk. This alignment brings two benefits. First, the trained model enjoys better abilities to generalize across scenarios with unseen cameras as well as transfer across multiple training sets. Second, we can rely on intra-camera annotations, which have been undervalued before due to the lack of cross-camera information, to achieve competitive ReID performance. Experiments on a wide range of ReID tasks demonstrate the effectiveness of our approach. The code is available at https://github.com/automan000/Camera-based-Person-ReID.

cs.CV↗

arXiv2020

Robust Multiscale Identification of Apparent Elastic Properties at Mesoscale for Random Heterogeneous Materials with Multiscale Field Measurements

The aim of this work is to efficiently and robustly solve the statistical inverse problem related to the identification of the elastic properties at both macroscopic and mesoscopic scales of heterogeneous anisotropic materials with a complex microstructure that usually cannot be properly described in terms of their mechanical constituents at microscale. Within the context of linear elasticity theory, the apparent elasticity tensor field at a given mesoscale is modeled by a prior non-Gaussian tensor-valued random field. A general methodology using multiscale displacement field measurements simultaneously made at both macroscale and mesoscale has been recently proposed for the identification the hyperparameters of such a prior stochastic model by solving a multiscale statistical inverse problem using a stochastic computational model and some information from displacement fields at both macroscale and mesoscale. This paper contributes to the improvement of the computational efficiency, accuracy and robustness of such a method by introducing (i) a mesoscopic numerical indicator related to the spatial correlation length(s) of kinematic fields, allowing the time-consuming global optimization algorithm (genetic algorithm) used in a previous work to be replaced with a more efficient algorithm and (ii) an ad hoc stochastic representation of the hyperparameters involved in the prior stochastic model in order to enhance both the robustness and the precision of the statistical inverse identification method. Finally, the proposed improved method is first validated on in silico materials within the framework of 2D plane stress and 3D linear elasticity (using multiscale simulated data obtained through numerical computations) and then exemplified on a real heterogeneous biological material (beef cortical bone) within the framework of 2D plane stress linear elasticity (using multiscale experimental data obtained through mechanical testing monitored by digital image correlation).

physics.class-ph↗

arXiv2020

Rethinking Privacy Preserving Deep Learning: How to Evaluate and Thwart Privacy Attacks

This paper investigates capabilities of Privacy-Preserving Deep Learning (PPDL) mechanisms against various forms of privacy attacks. First, we propose to quantitatively measure the trade-off between model accuracy and privacy losses incurred by reconstruction, tracing and membership attacks. Second, we formulate reconstruction attacks as solving a noisy system of linear equations, and prove that attacks are guaranteed to be defeated if condition (2) is unfulfilled. Third, based on theoretical analysis, a novel Secret Polarization Network (SPN) is proposed to thwart privacy attacks, which pose serious challenges to existing PPDL methods. Extensive experiments showed that model accuracies are improved on average by 5-20% compared with baseline mechanisms, in regimes where data privacy are satisfactorily protected.

cs.LG↗

arXiv2020

Interactive Rainbow Score: A Visual-centered Multimodal Flute Tutoring System

Learning to play an instrument is intrinsically multimodal, and we have seen a trend of applying visual and haptic feedback in music games and computer-aided music tutoring systems. However, most current systems are still designed to master individual pieces of music; it is unclear how well the learned skills can be generalized to new pieces. We aim to explore this question. In this study, we contribute Interactive Rainbow Score, an interactive visual system to boost the learning of sight-playing, the general musical skill to read music and map the visual representations to performance motions. The key design of Interactive Rainbow Score is to associate pitches (and the corresponding motions) with colored notation and further strengthen such association via real-time interactions. Quantitative results show that the interactive feature on average increases the learning efficiency by 31.1%. Further analysis indicates that it is critical to apply the interaction in the early period of learning.

cs.HC↗

arXiv2019

Neural Networks Weights Quantization: Target None-retraining Ternary (TNT)

Quantization of weights of deep neural networks (DNN) has proven to be an effective solution for the purpose of implementing DNNs on edge devices such as mobiles, ASICs and FPGAs, because they have no sufficient resources to support computation involving millions of high precision weights and multiply-accumulate operations. This paper proposes a novel method to compress vectors of high precision weights of DNNs to ternary vectors, namely a cosine similarity based target non-retraining ternary (TNT) compression method. Our method leverages cosine similarity instead of Euclidean distances as commonly used in the literature and succeeds in reducing the size of the search space to find optimal ternary vectors from 3N to N, where N is the dimension of target vectors. As a result, the computational complexity for TNT to find theoretically optimal ternary vectors is only O(N log(N)). Moreover, our experiments show that, when we ternarize models of DNN with high precision parameters, the obtained quantized models can exhibit sufficiently high accuracy so that re-training models is not necessary.

cs.LG↗

arXiv2019

Background Segmentation for Vehicle Re-Identification

Vehicle re-identification (Re-ID) is very important in intelligent transportation and video surveillance.Prior works focus on extracting discriminative features from visual appearance of vehicles or using visual-spatio-temporal information.However, background interference in vehicle re-identification have not been explored.In the actual large-scale spatio-temporal scenes, the same vehicle usually appears in different backgrounds while different vehicles might appear in the same background, which will seriously affect the re-identification performance. To the best of our knowledge, this paper is the first to consider the background interference problem in vehicle re-identification. We construct a vehicle segmentation dataset and develop a vehicle Re-ID framework with a background interference removal (BIR) mechanism to improve the vehicle Re-ID performance as well as robustness against complex background in large-scale spatio-temporal scenes. Extensive experiments demonstrate the effectiveness of our proposed framework, with an average 9% gain on mAP over state-of-the-art vehicle Re-ID algorithms.

cs.CV↗

arXiv2019

Single Camera Training for Person Re-identification

Person re-identification (ReID) aims at finding the same person in different cameras. Training such systems usually requires a large amount of cross-camera pedestrians to be annotated from surveillance videos, which is labor-consuming especially when the number of cameras is large. Differently, this paper investigates ReID in an unexplored single-camera-training (SCT) setting, where each person in the training set appears in only one camera. To the best of our knowledge, this setting was never studied before. SCT enjoys the advantage of low-cost data collection and annotation, and thus eases ReID systems to be trained in a brand new environment. However, it raises major challenges due to the lack of cross-camera person occurrences, which conventional approaches heavily rely on to extract discriminative features. The key to dealing with the challenges in the SCT setting lies in designing an effective mechanism to complement cross-camera annotation. We start with a regular deep network for feature extraction, upon which we propose a novel loss function named multi-camera negative loss (MCNL). This is a metric learning loss motivated by probability, suggesting that in a multi-camera system, one image is more likely to be closer to the most similar negative sample in other cameras than to the most similar negative sample in the same camera. In experiments, MCNL significantly boosts ReID accuracy in the SCT setting, which paves the way of fast deployment of ReID systems with good performance on new target scenes.

cs.CV↗

arXiv2019

Fully Distributed Packet Scheduling Framework for Handling Disturbances in Lossy Real-Time Wireless Networks

Along with the rapid growth of Industrial Internet-of-Things (IIoT) applications and their penetration into many industry sectors, real-time wireless networks (RTWNs) have been playing a more critical role in providing real-time, reliable and secure communication services for such applications. A key challenge in RTWN management is how to ensure real-time Quality of Services (QoS) especially in the presence of unexpected disturbances and lossy wireless links. Most prior work takes centralized approaches for handling disturbances, which are slow and subject to single-point failure, and do not scale. To overcome these drawbacks, this paper presents a fully distributed packet scheduling framework called FD-PaS. FD-PaS aims to provide guaranteed fast response to unexpected disturbances while achieving minimum performance degradation for meeting the timing and reliability requirements of all critical tasks. To combat the scalability challenge, FD-PaS incorporates several key advances in both algorithm design and data link layer protocol design to enable individual nodes to make on-line decisions locally without any centralized control. Our extensive simulation and testbed results have validated the correctness of the FD-PaS design and demonstrated its effectiveness in providing fast response for handling disturbances while ensuring the designated QoS requirements.

cs.NI↗

arXiv2018

Synergistic Drug Combination Prediction by Integrating Multi-omics Data in Deep Learning Models

Drug resistance is still a major challenge in cancer therapy. Drug combination is expected to overcome drug resistance. However, the number of possible drug combinations is enormous, and thus it is infeasible to experimentally screen all effective drug combinations considering the limited resources. Therefore, computational models to predict and prioritize effective drug combinations is important for combinatory therapy discovery in cancer. In this study, we proposed a novel deep learning model, AuDNNsynergy, to prediction drug combinations by integrating multi-omics data and chemical structure data. In specific, three autoencoders were trained using the gene expression, copy number and genetic mutation data of all tumor samples from The Cancer Genome Atlas. Then the physicochemical properties of drugs combined with the output of the three autoencoders, characterizing the individual cancer cell-lines, were used as the input of a deep neural network that predicts the synergy value of given pair-wise drug combinations against the specific cancer cell-lines. The comparison results showed the proposed AuDNNsynergy model outperforms four state-of-art approaches, namely DeepSynergy, Gradient Boosting Machines, Random Forests, and Elastic Nets. Moreover, we conducted the interpretation analysis of the deep learning model to investigate potential vital genetic predictors and the underlying mechanism of synergistic drug combinations on specific cancer cell-lines.

q-bio.GN↗

arXiv2018

Critical Multipliers in Semidefinite Programming

It was proved in [14] that the existence of a noncritical multiplier for a (smooth) nonlinear programming problem is equivalent to an error bound condition for the Karush-Kuhn-Thcker (KKT) system without any assumptions. This paper investigates whether this result still holds true for a (smooth) nonlinear semidefinite programming (SDP) problem. We first introduce the notion of critical and noncritical multipliers for a SDP problem and obtain their complete characterizations in terms of the problem data. We prove for the SDP problem, the noncriticality property can be derived from the error bound condition for the KKT system without any assumptions, and this fact is revealed by some simple examples. Besides we give an appropriate second-order sufficient optimality condition characterizing noncriticality explicitly. We propose a set of assumptions from which the error bound condition for the KKT system can be derived from the noncriticality property. Finally we establish a new error bound for x-part, which is expressed by both perturbation and the multiplier estimation.

math.OC↗

arXiv2016

Predict genome-scale fluxes based solely on enzyme abundance by a novel Hyper-Cube Shrink Algorithm

One of the long-expected goals of genome-scale metabolic modeling is to evaluate the influence of the perturbed enzymes to the flux distribution. Both ordinary differential equation (ODE) models and the constraint-based models, like Flux balance analysis (FBA), lack of the room of performing metabolic control analysis (MCA) for large-scale networks. In this study, we developed a Hyper-Cube Shrink Algorithm (HCSA) to incorporate the enzymatic properties to the FBA model by introducing a pseudo reaction constrained by enzymatic parameters. Our algorithm was able to handle not only prediction of knockout strains but also strains with quantitative adjustment of expression level or activity. We first demonstrate the concept by applying HCSA to a simplest three-node network. Then we validate its prediction by comparing with ODE and with a synthetic network in Saccharomyces cerevisiae producing voilacein and analogues. Finally we show its capability of predicting the flux distribution in genome-scale networks by applying it to the sporulation in yeast.

q-bio.MN↗