arXiv subjects

Rui Li

Publications and source records attributed to Rui Li.

At least 451 records · Page 25Linked to original sources

arXiv2021

Imitating Targets from all sides: An Unsupervised Transfer Learning method for Person Re-identification

Person re-identification (Re-ID) models usually show a limited performance when they are trained on one dataset and tested on another dataset due to the inter-dataset bias (e.g. completely different identities and backgrounds) and the intra-dataset difference (e.g. camera invariance). In terms of this issue, given a labelled source training set and an unlabelled target training set, we propose an unsupervised transfer learning method characterized by 1) bridging inter-dataset bias and intra-dataset difference via a proposed ImitateModel simultaneously; 2) regarding the unsupervised person Re-ID problem as a semi-supervised learning problem formulated by a dual classification loss to learn a discriminative representation across domains; 3) exploiting the underlying commonality across different domains from the class-style space to improve the generalization ability of re-ID models. Extensive experiments are conducted on two widely employed benchmarks, including Market-1501 and DukeMTMC-reID, and experimental results demonstrate that the proposed method can achieve a competitive performance against other state-of-the-art unsupervised Re-ID approaches.

cs.CV↗

arXiv2021

A unified Neural Network Approach to E-CommerceRelevance Learning

Result relevance scoring is critical to e-commerce search user experience. Traditional information retrieval methods focus on keyword matching and hand-crafted or counting-based numeric features, with limited understanding of item semantic relevance. We describe a highly-scalable feed-forward neural model to provide relevance score for (query, item) pairs, using only user query and item title as features, and both user click feedback as well as limited human ratings as labels. Several general enhancements were applied to further optimize eval/test metrics, including Siamese pairwise architecture, random batch negative co-training, and point-wise fine-tuning. We found significant improvement over GBDT baseline as well as several off-the-shelf deep-learning baselines on an independently constructed ratings dataset. The GBDT model relies on 10 times more features. We also present metrics for select subset combinations of techniques mentioned above.

cs.IR↗

arXiv2021

Quantized transport induced by topology transfer between coupled one-dimensional lattice systems

We show that a topological pump in a one-dimensional (1D) insulator can induce a strictly quantized transport in an auxiliary chain of non-interacting fermions weakly coupled to the first. The transported charge is determined by an integer topological invariant of the ficticious Hamiltonian of the insulator, given by the covariance matrix of single-particle correlations. If the original system consists of non-interacting fermions, this number is identical to the TKNN (Thouless, Kohmoto, Nightinghale, den Nijs) invariant of the original system and thus the coupling induces a transfer of topology to the auxiliary chain. When extended to particles with interactions, for which the TKNN number does not exist, the transported charge in the auxiliary chain defines a topological invariant for the interacting system. In certain cases this invariant agrees with the many-body generalization of the TKNN number introduced by Niu, Thouless, and Wu (NTW). We illustrate the topology transfer to the auxiliary system for the Rice-Mele model of non-interacting fermions at half filling and the extended superlattice Bose-Hubbard model at quarter filling. In the latter case the induced charge pump is fractional.

cond-mat.quant-gas↗

arXiv2021

From Semantic Retrieval to Pairwise Ranking: Applying Deep Learning in E-commerce Search

We introduce deep learning models to the two most important stages in product search at JD.com, one of the largest e-commerce platforms in the world. Specifically, we outline the design of a deep learning system that retrieves semantically relevant items to a query within milliseconds, and a pairwise deep re-ranking system, which learns subtle user preferences. Compared to traditional search systems, the proposed approaches are better at semantic retrieval and personalized ranking, achieving significant improvements.

cs.IR↗

arXiv2021

Adaptive Noise Injection: A Structure-Expanding Regularization for RNN

The vanilla LSTM has become one of the most potential architectures in word-level language modeling, like other recurrent neural networks, overfitting is always a key barrier for its effectiveness. The existing noise-injected regularizations introduce the random noises of fixation intensity, which inhibits the learning of the RNN throughout the training process. In this paper, we propose a new structure-expanding regularization method called Adjective Noise Injection (ANI), which considers the output of an extra RNN branch as a kind of adaptive noises and injects it into the main-branch RNN output. Due to the adaptive noises can be improved as the training processes, its negative effects can be weakened and even transformed into a positive effect to further improve the expressiveness of the main-branch RNN. As a result, ANI can regularize the RNN in the early stage of training and further promoting its training performance in the later stage. We conduct experiments on three widely-used corpora: PTB, WT2, and WT103, whose results verify both the regularization and promoting the training performance functions of ANI. Furthermore, we design a series simulation experiments to explore the reasons that may lead to the regularization effect of ANI, and we find that in training process, the robustness against the parameter update errors can be strengthened when the LSTM equipped with ANI.

cs.CL↗

arXiv2021

Searching spin-mass interaction using a diamagnetic levitated magnetic resonance force sensor

Axion-like particles (ALPs) are predicted to mediate exotic interactions between spin and mass. We propose an ALP-searching experiment based on the levitated micromechanical oscillator, which is one of the most sensitive sensors for spin-mass forces at a short distance. The proposed experiment tests the spin-mass resonant interaction between the polarized electron spins and a diamagnetically levitated microsphere. By periodically flipping the electron spins, the contamination from nonresonant background forces can be eliminated. The levitated microoscillator can prospectively enhance the sensitivity by nearly $10^3$ times over current experiments for ALPs with mass in the range 4 meV to 0.4 eV.

quant-ph↗

arXiv2021

Analytical Characterization and Design Space Exploration for Optimization of CNNs

Moving data through the memory hierarchy is a fundamental bottleneck that can limit the performance of core algorithms of machine learning, such as convolutional neural networks (CNNs). Loop-level optimization, including loop tiling and loop permutation, are fundamental transformations to reduce data movement. However, the search space for finding the best loop-level optimization configuration is explosively large. This paper develops an analytical modeling approach for finding the best loop-level optimization configuration for CNNs on multi-core CPUs. Experimental evaluation shows that this approach achieves comparable or better performance than state-of-the-art libraries and auto-tuning based optimizers for CNNs.

cs.LG↗

arXiv2021

Driver Behavior Recognition via Interwoven Deep Convolutional Neural Nets with Multi-stream Inputs

Understanding driver activity is vital for in-vehicle systems that aim to reduce the incidence of car accidents rooted in cognitive distraction. Automating real-time behavior recognition while ensuring actions classification with high accuracy is however challenging, given the multitude of circumstances surrounding drivers, the unique traits of individuals, and the computational constraints imposed by in-vehicle embedded platforms. Prior work fails to jointly meet these runtime/accuracy requirements and mostly rely on a single sensing modality, which in turn can be a single point of failure. In this paper, we harness the exceptional feature extraction abilities of deep learning and propose a dedicated Interwoven Deep Convolutional Neural Network (InterCNN) architecture to tackle the problem of accurate classification of driver behaviors in real-time. The proposed solution exploits information from multi-stream inputs, i.e., in-vehicle cameras with different fields of view and optical flows computed based on recorded images, and merges through multiple fusion layers abstract features that it extracts. This builds a tight ensembling system, which significantly improves the robustness of the model. In addition, we introduce a temporal voting scheme based on historical inference instances, to enhance the classification accuracy. Experiments conducted with a dataset that we collect in a mock-up car environment demonstrate that the proposed InterCNN with MobileNet convolutional blocks can classify 9 different behaviors with 73.97% accuracy, and 5 'aggregated' behaviors with 81.66% accuracy. We further show that our architecture is highly computationally efficient, as it performs inferences within 15ms, which satisfies the real-time constraints of intelligent cars. Nevertheless, our InterCNN is robust to lossy input, as the classification remains accurate when two input streams are occluded.

cs.CV↗

arXiv2021

Learning Depth via Leveraging Semantics: Self-supervised Monocular Depth Estimation with Both Implicit and Explicit Semantic Guidance

Self-supervised depth estimation has made a great success in learning depth from unlabeled image sequences. While the mappings between image and pixel-wise depth are well-studied in current methods, the correlation between image, depth and scene semantics, however, is less considered. This hinders the network to better understand the real geometry of the scene, since the contextual clues, contribute not only the latent representations of scene depth, but also the straight constraints for depth map. In this paper, we leverage the two benefits by proposing the implicit and explicit semantic guidance for accurate self-supervised depth estimation. We propose a Semantic-aware Spatial Feature Alignment (SSFA) scheme to effectively align implicit semantic features with depth features for scene-aware depth estimation. We also propose a semantic-guided ranking loss to explicitly constrain the estimated depth maps to be consistent with real scene contextual properties. Both semantic label noise and prediction uncertainty is considered to yield reliable depth supervisions. Extensive experimental results show that our method produces high quality depth maps which are consistently superior either on complex scenes or diverse semantic categories, and outperforms the state-of-the-art methods by a significant margin.

cs.CV↗

arXiv2021

ABCNet: Attentive Bilateral Contextual Network for Efficient Semantic Segmentation of Fine-Resolution Remote Sensing Images

Semantic segmentation of remotely sensed images plays a crucial role in precision agriculture, environmental protection, and economic assessment. In recent years, substantial fine-resolution remote sensing images are available for semantic segmentation. However, due to the complicated information caused by the increased spatial resolution, state-of-the-art deep learning algorithms normally utilize complex network architectures for segmentation, which usually incurs high computational complexity. Specifically, the high-caliber performance of the convolutional neural network (CNN) heavily relies on fine-grained spatial details (fine resolution) and sufficient contextual information (large receptive fields), both of which trigger high computational costs. This crucially impedes their practicability and availability in real-world scenarios that require real-time processing. In this paper, we propose an Attentive Bilateral Contextual Network (ABCNet), a convolutional neural network (CNN) with double branches, with prominently lower computational consumptions compared to the cutting-edge algorithms, while maintaining a competitive accuracy. Code is available at https://github.com/lironui/ABCNet.

cs.CV↗

arXiv2021

Non-uniform Motion Deblurring with Blurry Component Divided Guidance

Blind image deblurring is a fundamental and challenging computer vision problem, which aims to recover both the blur kernel and the latent sharp image from only a blurry observation. Despite the superiority of deep learning methods in image deblurring have displayed, there still exists major challenge with various non-uniform motion blur. Previous methods simply take all the image features as the input to the decoder, which handles different degrees (e.g. large blur, small blur) simultaneously, leading to challenges for sharp image generation. To tackle the above problems, we present a deep two-branch network to deal with blurry images via a component divided module, which divides an image into two components based on the representation of blurry degree. Specifically, two component attentive blocks are employed to learn attention maps to exploit useful deblurring feature representations on both large and small blurry regions. Then, the blur-aware features are fed into two-branch reconstruction decoders respectively. In addition, a new feature fusion mechanism, orientation-based feature fusion, is proposed to merge sharp features of the two branches. Both qualitative and quantitative experimental results show that our method performs favorably against the state-of-the-art approaches.

cs.CV↗

arXiv2021

Unfolding the Neutron Spectra from a Water-Pumping-Injection Multi-layered Concentric Sphere Neutron Spectrometer Using a Self-Adaptive Differential Evolution Algorithm

A self-adaptive differential evolution neutron spectrum unfolding algorithm (SDENUA) was established in this paper to unfold the neutron spectra obtained from a Water-pumping-injection Multi-layered concentric sphere Neutron Spectrometer (WMNS). Specifically, the neutron fluence bounds were estimated to accelerate the algorithm convergence, the minimum error between the optimal solution and the input neutron counts with relative uncertainties was limited to 10-6 to avoid useless calculation. Furthermore, the crossover probability and scaling factor were controlled self-adaptively. FLUKA Monte Carlo was used to simulate the readings of the WMNS under (1) a spectrum of Cf-252 and (2) its spectrum after being moderated, (3) a spectrum used for BNCT, and (4) a reactor spectrum, and the measured neutron counts unfolded by using the SDENUA. The uncertainties of the measured neutron count and the response matrix are considered in the SDENUA, which does not require complex parameter tuning and the priori default spectrum. Results indicate that the solutions of the SDENUA are more in agreement with the IAEA spectra than that of the MAXED and GRAVEL in UMG 3.1, and the errors of the final results calculated by SDENUA are under 12%. The established SDENUA has potential applications for unfolding spectra from the WMNS.

physics.ins-det↗

arXiv2021

Generalize Ultrasound Image Segmentation via Instant and Plug & Play Style Transfer

Deep segmentation models that generalize to images with unknown appearance are important for real-world medical image analysis. Retraining models leads to high latency and complex pipelines, which are impractical in clinical settings. The situation becomes more severe for ultrasound image analysis because of their large appearance shifts. In this paper, we propose a novel method for robust segmentation under unknown appearance shifts. Our contribution is three-fold. First, we advance a one-stage plug-and-play solution by embedding hierarchical style transfer units into a segmentation architecture. Our solution can remove appearance shifts and perform segmentation simultaneously. Second, we adopt Dynamic Instance Normalization to conduct precise and dynamic style transfer in a learnable manner, rather than previously fixed style normalization. Third, our solution is fast and lightweight for routine clinical adoption. Given 400*400 image input, our solution only needs an additional 0.2ms and 1.92M FLOPs to handle appearance shifts compared to the baseline pipeline. Extensive experiments are conducted on a large dataset from three vendors demonstrate our proposed method enhances the robustness of deep segmentation models.

eess.IV↗

arXiv2021

Research on Fitness Function of Two Evolution Algorithms Used for Neutron Spectrum Unfolding

When evolution algorithms are used to unfold the neutron energy spectrum, fitness function design is an important fundamental work for evaluating the quality of the solution, but it has not attracted much attention. In this work, we investigated the performance of eight fitness functions attached to the genetic algorithm (GA) and the differential evolution algorithm (DEA) used for unfolding four neutron spectra selected from the IAEA 403 report. Experiments show that the fitness functions with a maximum in the GA can limit the ability of the population to percept the fitness change, but the ability can be made up in the DEA. The fitness function with a feature penalty term helps to improve the performance of solutions, and the fitness function using the standard deviation and the Chi-squared result shows the balance between the algorithm and the spectra. The results also show that the DEA has good potential for neutron energy spectrum unfolding. The purposes of this work are to provide evidence for structuring and modifying the fitness functions and to suggest some genetic operations that should receive attention when using the fitness function to unfold neutron spectra.

cs.NE↗

arXiv2021

VIS30K: A Collection of Figures and Tables from IEEE Visualization Conference Publications

We present the VIS30K dataset, a collection of 29,689 images that represents 30 years of figures and tables from each track of the IEEE Visualization conference series (Vis, SciVis, InfoVis, VAST). VIS30K's comprehensive coverage of the scientific literature in visualization not only reflects the progress of the field but also enables researchers to study the evolution of the state-of-the-art and to find relevant work based on graphical content. We describe the dataset and our semi-automatic collection process, which couples convolutional neural networks (CNN) with curation. Extracting figures and tables semi-automatically allows us to verify that no images are overlooked or extracted erroneously. To improve quality further, we engaged in a peer-search process for high-quality figures from early IEEE Visualization papers. With the resulting data, we also contribute VISImageNavigator (VIN, visimagenavigator.github.io), a web-based tool that facilitates searching and exploring VIS30K by author names, paper keywords, title and abstract, and years.

cs.CV↗

arXiv2020

Aligning free surface properties in time-varying hydraulic jumps

Hydraulic jumps occur commonly in natural channels and energy dissipation systems of hydraulic structures in the violent transition from supercritical to subcritical flows. They are characterised by large flow aeration, high turbulence and strong fluctuations of the free surface and the jump toe. For free surface measurements, fast-sampling, fixed-point instruments such as acoustic displacement meters (ADMs) and wire gauges (WGs) are commonly used, while LIDAR technology is a relatively new method for recording instantaneous free surface motions of aerated flows. While each of these instruments has been shown previously to provide reasonable results for basic and advanced free surface properties, differences between instruments and experiments remain unexplained. To systematically analyse these differences, simultaneous laboratory experiments of aerated hydraulic jumps were conducted. Good agreement between the three instruments was obtained for basic free surface properties including elevations, fluctuations, skewness, kurtosis, and frequencies, as well as advanced free surface properties such as integral time and length scales. These new results indicate that any of these instruments can be used for the recording of free surface properties albeit the integration limit for free surface scales must be considered. A key finding of this research was that differences between repeated experiments as well as previous studies were observed when using the visual jump toe for alignment. However, this bias could be resolved by using the mean jump toe location recorded with the LIDAR. Therefore, future studies should simultaneously measure the instantaneous jump toe to provide more consistent results across studies.

physics.flu-dyn↗

arXiv2020

Multi-Head Linear Attention Generative Adversarial Network for Thin Cloud Removal

In remote sensing images, the existence of the thin cloud is an inevitable and ubiquitous phenomenon that crucially reduces the quality of imageries and limits the scenarios of application. Therefore, thin cloud removal is an indispensable procedure to enhance the utilization of remote sensing images. Generally, even though contaminated by thin clouds, the pixels still retain more or less surface information. Hence, different from thick cloud removal, thin cloud removal algorithms normally concentrate on inhibiting the cloud influence rather than substituting the cloud-contaminated pixels. Meanwhile, considering the surface features obscured by the cloud are usually similar to adjacent areas, the dependency between each pixel of the input is useful to reconstruct contaminated areas. In this paper, to make full use of the dependencies between pixels of the image, we propose a Multi-Head Linear Attention Generative Adversarial Network (MLAGAN) for Thin Cloud Removal. The MLA-GAN is based on the encoding-decoding framework consisting of multiple attention-based layers and deconvolutional layers. Compared with six deep learning-based thin cloud removal benchmarks, the experimental results on the RICE1 and RICE2 datasets demonstrate that the proposed framework MLA-GAN has dominant advantages in thin cloud removal.

cs.CV↗

arXiv2020

Semantic-Guided Representation Enhancement for Self-supervised Monocular Trained Depth Estimation

Self-supervised depth estimation has shown its great effectiveness in producing high quality depth maps given only image sequences as input. However, its performance usually drops when estimating on border areas or objects with thin structures due to the limited depth representation ability. In this paper, we address this problem by proposing a semantic-guided depth representation enhancement method, which promotes both local and global depth feature representations by leveraging rich contextual information. In stead of a single depth network as used in conventional paradigms, we propose an extra semantic segmentation branch to offer extra contextual features for depth estimation. Based on this framework, we enhance the local feature representation by sampling and feeding the point-based features that locate on the semantic edges to an individual Semantic-guided Edge Enhancement module (SEEM), which is specifically designed for promoting depth estimation on the challenging semantic borders. Then, we improve the global feature representation by proposing a semantic-guided multi-level attention mechanism, which enhances the semantic and depth features by exploring pixel-wise correlations in the multi-level depth decoding scheme. Extensive experiments validate the distinct superiority of our method in capturing highly accurate depth on the challenging image areas such as semantic category borders and thin objects. Both quantitative and qualitative experiments on KITTI show that our method outperforms the state-of-the-art methods.

cs.CV↗