arXiv subjects

Yu Li

Publications and source records attributed to Yu Li.

At least 433 records · Page 24Linked to original sources

arXiv2022

Anisotropic gap structure and sign reversal symmetry in monolayer Fe(Se,Te)

The iron-based superconductors are an ideal platform to reveal the enigma of the unconventional superconductivity and potential topological superconductivity. Among them, the monolayer Fe(Se,Te)/SrTiO3(001), which is proposed to be topological nontrivial, shows interface-enhanced high-temperature superconductivity in the two dimensional limit. However, the experimental studies on the superconducting pairing mechanism of monolayer Fe(Se,Te) films are still limited. Here, by measuring quasiparticle interference in monolayer Fe(Se,Te)/SrTiO3(001), we report the observation of the anisotropic structure of the large superconducting gap and the sign change of the superconducting gap on different electron pockets. The results are well consistent with the 'bonding-antibonding' s+- wave pairing symmetry driven by spin fluctuations in conjunction with spin-orbit coupling. Our work is of basic significance not only for a unified superconducting formalism in the iron-based superconductors, but also for understanding of topological superconductivity in high-temperature superconductors.

cond-mat.supr-con↗

arXiv2022

Multi-Class Anomaly Detection

We study anomaly detection for the case when the normal class consists of more than one object category. This is an obvious generalization of the standard one-class anomaly detection problem. However, we show that jointly using multiple one-class anomaly detectors to solve this problem yields poorer results as compared to training a single one-class anomaly detector on all normal object categories together. We further develop a new anomaly detector called DeepMAD that learns compact distinguishing features by exploiting the multiple normal objects categories. This algorithm achieves higher AUC values for different datasets compared to two top performing one-class algorithms that either are trained on each normal object category or jointly trained on all normal object categories combined. In addition to theoretical results we present empirical results using the CIFAR-10, fMNIST, CIFAR-100, and a new dataset we developed called RECYCLE.

cs.LG↗

arXiv2022

Learning Single Image Defocus Deblurring with Misaligned Training Pairs

By adopting popular pixel-wise loss, existing methods for defocus deblurring heavily rely on well aligned training image pairs. Although training pairs of ground-truth and blurry images are carefully collected, e.g., DPDD dataset, misalignment is inevitable between training pairs, making existing methods possibly suffer from deformation artifacts. In this paper, we propose a joint deblurring and reblurring learning (JDRL) framework for single image defocus deblurring with misaligned training pairs. Generally, JDRL consists of a deblurring module and a spatially invariant reblurring module, by which deblurred result can be adaptively supervised by ground-truth image to recover sharp textures while maintaining spatial consistency with the blurry image. First, in the deblurring module, a bi-directional optical flow-based deformation is introduced to tolerate spatial misalignment between deblurred and ground-truth images. Second, in the reblurring module, deblurred result is reblurred to be spatially aligned with blurry image, by predicting a set of isotropic blur kernels and weighting maps. Moreover, we establish a new single image defocus deblurring (SDD) dataset, further validating our JDRL and also benefiting future research. Our JDRL can be applied to boost defocus deblurring networks in terms of both quantitative metrics and visual quality on DPDD, RealDOF and our SDD datasets.

cs.CV↗

arXiv2022

Do we really need temporal convolutions in action segmentation?

Action classification has made great progress, but segmenting and recognizing actions from long untrimmed videos remains a challenging problem. Most state-of-the-art methods focus on designing temporal convolution-based models, but the inflexibility of temporal convolutions and the difficulties in modeling long-term temporal dependencies restrict the potential of these models. Transformer-based models with adaptable and sequence modeling capabilities have recently been used in various tasks. However, the lack of inductive bias and the inefficiency of handling long video sequences limit the application of Transformer in action segmentation. In this paper, we design a pure Transformer-based model without temporal convolutions by incorporating temporal sampling, called Temporal U-Transformer (TUT). The U-Transformer architecture reduces complexity while introducing an inductive bias that adjacent frames are more likely to belong to the same class, but the introduction of coarse resolutions results in the misclassification of boundaries. We observe that the similarity distribution between a boundary frame and its neighboring frames depends on whether the boundary frame is the start or end of an action segment. Therefore, we further propose a boundary-aware loss based on the distribution of similarity scores between frames from attention modules to enhance the ability to recognize boundaries. Extensive experiments show the effectiveness of our model.

cs.CV↗

arXiv2022

A Simulation Study of Passing Drivers' Responses to the Autonomous Truck-Mounted Attenuator System in Road Maintenance

The Autonomous Truck-Mounted Attenuator (ATMA) system is a lead-follower vehicle system based on autonomous driving and connected vehicle technologies. The lead truck performs maintenance tasks on the road, and the unmanned follower truck alerts passing vehicles about the moving work zone and protects workers and the equipment. While the ATMA has been under testing by transportation maintenance and operations agencies recently, a simulator-based testing capability is a supplement, especially if human subjects are involved. This paper aims to discover how passing drivers perceive, understand, and react to the ATMA system in road maintenance. With the driving simulator developed for this ATMA study, the paper performed a simulation study wherein a screen-based eye tracker collected sixteen subjects' gaze points and pupil diameters. Data analysis evidenced the change in subjects' visual attention patterns while passing the ATMA. On average, the ATMA starts to attract subjects' attention from 500 ft behind the follower truck. Most (87.50%) understood the follower truck's protection purpose, and many (60%) reasoned the association between the two trucks. Nevertheless, nearly half of the participants (43.75%) did not recognize that ATMA is a connected autonomous vehicle system. While all subjects safely changed lanes and attempted to pass the slow-moving ATMA, their inadequate understanding of the ATMA is a potential risk, like cutting into the ATAM. Results implied that transportation maintenance and operations agencies should consider this in establishing the deployment guidance.

cs.RO↗

arXiv2022

A Gaze Data-based Comparative Study to Build a Trustworthy Human-AI Collaboration in Crash Anticipation

Vehicles with a safety function for anticipating crashes in advance can enhance drivers' ability to avoid crashes. As dashboard cameras have become a low-cost sensor device accessible to almost every vehicle, deep neural networks for crash anticipation from a dashboard camera are receiving growing interest. However, drivers' trust in the Artificial Intelligence (AI)-enabled safety function is built on the validation of its safety enhancement toward zero deaths. This paper is motivated to establish a method that uses gaze data and corresponding measures to evaluate human drivers' ability to anticipate crashes. A laboratory experiment is designed and performed, wherein a screen-based eye tracker collects the gaze data of six volunteers while watching 100 driving videos that include both normal and crash scenarios. Statistical analyses of the experimental data show that, on average, drivers can anticipate a crash up to 2.61 seconds before it occurs in this pilot study. The chance that drivers have successfully anticipated crashes before they occur is 92.8%. A state-of-the-art AI model can anticipate crashes 1.02 seconds earlier than drivers on average. The study finds that crash-involving traffic agents in the driving videos can vary drivers' instant attention level, average attention level, and spatial attention distribution. This finding supports the development of a spatial-temporal attention mechanism for AI models to strengthen their ability to anticipate crashes. Results from the comparison also suggest the development of collaborative intelligence that keeps human-in-the-loop of AI models to further enhance the reliability of AI-enabled safety functions.

cs.HC↗

arXiv2022

A Multi-tasking Model of Speaker-Keyword Classification for Keeping Human in the Loop of Drone-assisted Inspection

Audio commands are a preferred communication medium to keep inspectors in the loop of civil infrastructure inspection performed by a semi-autonomous drone. To understand job-specific commands from a group of heterogeneous and dynamic inspectors, a model must be developed cost-effectively for the group and easily adapted when the group changes. This paper is motivated to build a multi-tasking deep learning model that possesses a Share-Split-Collaborate architecture. This architecture allows the two classification tasks to share the feature extractor and then split subject-specific and keyword-specific features intertwined in the extracted features through feature projection and collaborative training. A base model for a group of five authorized subjects is trained and tested on the inspection keyword dataset collected by this study. The model achieved a 95.3% or higher mean accuracy in classifying the keywords of any authorized inspectors. Its mean accuracy in speaker classification is 99.2%. Due to the richer keyword representations that the model learns from the pooled training data, adapting the base model to a new inspector requires only a little training data from that inspector, like five utterances per keyword. Using the speaker classification scores for inspector verification can achieve a success rate of at least 93.9% in verifying authorized inspectors and 76.1% in detecting unauthorized ones. Further, the paper demonstrates the applicability of the proposed model to larger-size groups on a public dataset. This paper provides a solution to addressing challenges facing AI-assisted human-robot interaction, including worker heterogeneity, worker dynamics, and job heterogeneity.

cs.SD↗

arXiv2022

Refine and Imitate: Reducing Repetition and Inconsistency in Persuasion Dialogues via Reinforcement Learning and Human Demonstration

Persuasion dialogue systems reflect the machine's ability to make strategic moves beyond verbal communication, and therefore differentiate themselves from task-oriented or open-domain dialogue systems and have their own unique values. However, the repetition and inconsistency problems still persist in dialogue response generation and could substantially impact user experience and impede the persuasion outcome. Besides, although reinforcement learning (RL) approaches have achieved big success in strategic tasks such as games, they require a sophisticated user simulator to provide real-time feedback to the dialogue system, which limits the application of RL on persuasion dialogues. To address these issues towards a better persuasion dialogue system, we apply RL to refine a language model baseline without user simulators, and distill sentence-level information about repetition, inconsistency, and task relevance through rewards. Moreover, to better accomplish the persuasion task, the model learns from human demonstration to imitate human persuasion behavior and selects the most persuasive responses. Experiments show that our model outperforms previous state-of-the-art dialogue models on both automatic metrics and human evaluation results on a donation persuasion task, and generates more diverse, consistent and persuasive conversations according to the user feedback.

cs.CL↗

arXiv2022

Robots-Dont-Cry: Understanding Falsely Anthropomorphic Utterances in Dialog Systems

Dialog systems are often designed or trained to output human-like responses. However, some responses may be impossible for a machine to truthfully say (e.g. "that movie made me cry"). Highly anthropomorphic responses might make users uncomfortable or implicitly deceive them into thinking they are interacting with a human. We collect human ratings on the feasibility of approximately 900 two-turn dialogs sampled from 9 diverse data sources. Ratings are for two hypothetical machine embodiments: a futuristic humanoid robot and a digital assistant. We find that for some data-sources commonly used to train dialog systems, 20-30% of utterances are not viewed as possible for a machine. Rating is marginally affected by machine embodiment. We explore qualitative and quantitative reasons for these ratings. Finally, we build classifiers and explore how modeling configuration might affect output permissibly, and discuss implications for building less falsely anthropomorphic dialog systems.

cs.CL↗

arXiv2022

Two-stage superconductivity in the Hatsugai-Kohmoto-BCS model

Superconductivity in strongly correlated electrons can emerge out from a normal state that is beyond the Landau's Fermi liquid paradigm, often dubbed as "non-Fermi liquid". While the theory for non-Fermi liquid is still not yet conclusive, a recent study on the exactly-solvable Hatsugai-Kohmoto (HK) model has suggested a non-Fermi liquid ground state whose Green's function resembles the Yang-Rice-Zhang ansatz for cuprates [P. W. Phillips, L. Yeo and E. W. Huang, Nat. Phys. $\bf{16}$, 1175 (2020)]. Similar to the effect of on-site Coulomb repulsion in the Hubbard model, the repulsive interaction in the HK model divides the momentum space into three parts: empty, single-occupied and double-occupied regions, that are separated from each other by two distinct Fermi surfaces. In the presence of an additional Bardeen-Cooper-Schrieffer (BCS)-type pairing interaction of a moderate strength, we show that the system exhibits a "two-stage superconductivity" feature as temperature decreases: a first-order superconducting transition occurs at a temperature $T_{\rm c}$ that is followed by a sudden increase of the superconducting order parameter at a lower temperature $T_{\rm c}^{\prime}<T_{\rm c}$. At the first stage, $T_{\rm c}^{\prime}<T<T_{\rm c}$, the pairing function arises and the entropy is released only in the vicinity of the two Fermi surfaces; while at the second stage, $T<T_{\rm c}^{\prime}$, the pairing function becomes significant and the entropy is further released in deep (single-occupied) region in the Fermi sea. The phase transitions are analyzed within the Ginzburg-Landau theory. Our work sheds new light on unconventional superconductivity in strongly correlated electrons.

cond-mat.supr-con↗

arXiv2022

Robust Human Matting via Semantic Guidance

Automatic human matting is highly desired for many real applications. We investigate recent human matting methods and show that common bad cases happen when semantic human segmentation fails. This indicates that semantic understanding is crucial for robust human matting. From this, we develop a fast yet accurate human matting framework, named Semantic Guided Human Matting (SGHM). It builds on a semantic human segmentation network and introduces a light-weight matting module with only marginal computational cost. Unlike previous works, our framework is data efficient, which requires a small amount of matting ground-truth to learn to estimate high quality object mattes. Our experiments show that trained with merely 200 matting images, our method can generalize well to real-world datasets, and outperform recent methods on multiple benchmarks, while remaining efficient. Considering the unbearable labeling cost of matting data and widely available segmentation data, our method becomes a practical and effective solution for the task of human matting. Source code is available at https://github.com/cxgincsu/SemanticGuidedHumanMatting.

cs.CV↗

arXiv2022

GTAV-NightRain: Photometric Realistic Large-scale Dataset for Night-time Rain Streak Removal

Rain is transparent, which reflects and refracts light in the scene to the camera. In outdoor vision, rain, especially rain streaks degrade visibility and therefore need to be removed. In existing rain streak removal datasets, although density, scale, direction and intensity have been considered, transparency is not fully taken into account. This problem is particularly serious in night scenes, where the appearance of rain largely depends on the interaction with scene illuminations and changes drastically on different positions within the image. This is problematic, because unrealistic dataset causes serious domain bias. In this paper, we propose GTAV-NightRain dataset, which is a large-scale synthetic night-time rain streak removal dataset. Unlike existing datasets, by using 3D computer graphic platform (namely GTA V), we are allowed to infer the three dimensional interaction between rain and illuminations, which insures the photometric realness. Current release of the dataset contains 12,860 HD rainy images and 1,286 corresponding HD ground truth images in diversified night scenes. A systematic benchmark and analysis are provided along with the dataset to inspire further research.

cs.CV↗

arXiv2022

ViTKD: Practical Guidelines for ViT feature knowledge distillation

Knowledge Distillation (KD) for Convolutional Neural Network (CNN) is extensively studied as a way to boost the performance of a small model. Recently, Vision Transformer (ViT) has achieved great success on many computer vision tasks and KD for ViT is also desired. However, besides the output logit-based KD, other feature-based KD methods for CNNs cannot be directly applied to ViT due to the huge structure gap. In this paper, we explore the way of feature-based distillation for ViT. Based on the nature of feature maps in ViT, we design a series of controlled experiments and derive three practical guidelines for ViT's feature distillation. Some of our findings are even opposite to the practices in the CNN era. Based on the three guidelines, we propose our feature-based method ViTKD which brings consistent and considerable improvement to the student. On ImageNet-1k, we boost DeiT-Tiny from 74.42% to 76.06%, DeiT-Small from 80.55% to 81.95%, and DeiT-Base from 81.76% to 83.46%. Moreover, ViTKD and the logit-based KD method are complementary and can be applied together directly. This combination can further improve the performance of the student. Specifically, the student DeiT-Tiny, Small, and Base achieve 77.78%, 83.59%, and 85.41%, respectively. The code is available at https://github.com/yzd-v/cls_KD.

cs.CV↗

arXiv2022

Multipole-fluctuation pairing mechanism of $d_{x^2-y^2}+ig$ superconductivity in Sr$_2$RuO$_4$

Despite of many experimental and theoretical efforts, the pairing symmetry of superconductivity in Sr$_2$RuO$_4$ remains undecided. The accidentally degenerate $d_{x^2-y^2}+ig$ is consistent with most current experiments and seems to be one of the most probable candidates, but we still lack a satisfactory theoretical mechanism for its appearance. Here we construct a phenomenological model combining realistic electronic band structures and all symmetry-allowed multipole fluctuations as potential pairing glues, and make a systematic survey of major pairing states within the Eliashberg framework. Our calculations show that $d_{x^2-y^2}+ig$ can arise naturally from the interplay of antiferromagnetic, ferromagnetic, and electric multipole fluctuations whose coexistence is manifested in previous experiments and calculations. Our work provides a physically reasonable basis supporting the possibility of $d_{x^2-y^2}+ig$ pairing in superconducting Sr$_2$RuO$_4$.

cond-mat.supr-con↗

arXiv2022

Rethinking Knowledge Distillation via Cross-Entropy

Knowledge Distillation (KD) has developed extensively and boosted various tasks. The classical KD method adds the KD loss to the original cross-entropy (CE) loss. We try to decompose the KD loss to explore its relation with the CE loss. Surprisingly, we find it can be regarded as a combination of the CE loss and an extra loss which has the identical form as the CE loss. However, we notice the extra loss forces the student's relative probability to learn the teacher's absolute probability. Moreover, the sum of the two probabilities is different, making it hard to optimize. To address this issue, we revise the formulation and propose a distributed loss. In addition, we utilize teachers' target output as the soft target, proposing the soft loss. Combining the soft loss and the distributed loss, we propose a new KD loss (NKD). Furthermore, we smooth students' target output to treat it as the soft target for training without teachers and propose a teacher-free new KD loss (tf-NKD). Our method achieves state-of-the-art performance on CIFAR-100 and ImageNet. For example, with ResNet-34 as the teacher, we boost the ImageNet Top-1 accuracy of ResNet18 from 69.90% to 71.96%. In training without teachers, MobileNet, ResNet-18 and SwinTransformer-Tiny achieve 70.04%, 70.76%, and 81.48%, which are 0.83%, 0.86%, and 0.30% higher than the baseline, respectively. The code is available at https://github.com/yzd-v/cls_KD.

cs.CV↗

arXiv2022

Genome-wide nucleotide-resolution model of single-strand break site reveals species evolutionary hierarchy

Single-strand breaks (SSBs) are the major DNA damage in the genome arising spontaneously as the outcome of genotoxins and intermediates of DNA transactions. SSBs play a crucial role in various biological processes and show a non-random distribution in the genome. Several SSB detection approaches such as S1 END-seq and SSiNGLe-ILM emerged to characterize the genomic landscape of SSB with nucleotide resolution. However, these sequencing-based methods are costly and unfeasible for large-scale analysis of diverse species. Thus, we proposed the first computational approach, SSBlazer, which is an explainable and scalable deep learning framework for genome-wide nucleotide-resolution SSB site prediction. We demonstrated that SSBlazer can accurately predict SSB sites and sufficiently alleviate false positives by constructing an imbalanced dataset to simulate the realistic SSB distribution. The model interpretation analysis reveals that SSBlazer captures the pattern of individual CpG in genomic context and the motif of TGCC in the center region as critical features. Besides, SSBlazer is a lightweight model with robust cross-species generalization ability in the cross-species evaluation, which enables the large-scale genome-wide application in diverse species. Strikingly, the putative SSB genomic landscapes of 216 vertebrates reveal a negative correlation between SSB frequency and evolutionary hierarchy, suggesting that the genome tends to be integrity during evolution.

q-bio.GN↗

arXiv2022

Towards Vivid and Diverse Image Colorization with Generative Color Prior

Colorization has attracted increasing interest in recent years. Classic reference-based methods usually rely on external color images for plausible results. A large image database or online search engine is inevitably required for retrieving such exemplars. Recent deep-learning-based methods could automatically colorize images at a low cost. However, unsatisfactory artifacts and incoherent colors are always accompanied. In this work, we propose GCP-Colorization that leverages the rich and diverse color priors encapsulated in a pretrained Generative Adversarial Networks (GAN) for automatic colorization. Specifically, we first "retrieve" matched features (similar to exemplars) via a GAN encoder and then incorporate these features into the colorization process with feature modulations. Thanks to the powerful generative color prior (GCP) and delicate designs, our GCP-Colorization could produce vivid colors with a single forward pass. Moreover, it is highly convenient to obtain diverse results by modifying GAN latent codes. GCP-Colorization also inherits the merit of interpretable controls of GANs and could attain controllable and smooth transitions by walking through GAN latent space. Extensive experiments and user studies demonstrate that GCP-Colorization achieves superior performance than previous works. Codes are available at https://github.com/ToTheBeginning/GCP-Colorization.

cs.CV↗

arXiv2022

Interpretable RNA Foundation Model from Unannotated Data for Highly Accurate RNA Structure and Function Predictions

Non-coding RNA structure and function are essential to understanding various biological processes, such as cell signaling, gene expression, and post-transcriptional regulations. These are all among the core problems in the RNA field. With the rapid growth of sequencing technology, we have accumulated a massive amount of unannotated RNA sequences. On the other hand, expensive experimental observatory results in only limited numbers of annotated data and 3D structures. Hence, it is still challenging to design computational methods for predicting their structures and functions. The lack of annotated data and systematic study causes inferior performance. To resolve the issue, we propose a novel RNA foundation model (RNA-FM) to take advantage of all the 23 million non-coding RNA sequences through self-supervised learning. Within this approach, we discover that the pre-trained RNA-FM could infer sequential and evolutionary information of non-coding RNAs without using any labels. Furthermore, we demonstrate RNA-FM's effectiveness by applying it to the downstream secondary/3D structure prediction, SARS-CoV-2 genome structure and evolution prediction, protein-RNA binding preference modeling, and gene expression regulation modeling. The comprehensive experiments show that the proposed method improves the RNA structural and functional modelling results significantly and consistently. Despite only being trained with unlabelled data, RNA-FM can serve as the foundational model for the field.

q-bio.QM↗