Searcharxiv⌕ Search

arXiv subjects

Jing Liu

Publications and source records attributed to Jing Liu.

At least 433 records · Page 24Linked to original sources
arXiv2023

Learning Causality-inspired Representation Consistency for Video Anomaly Detection

Video anomaly detection is an essential yet challenging task in the multimedia community, with promising applications in smart cities and secure communities. Existing methods attempt to learn abstract representations of regular events with statistical dependence to model the endogenous normality, which discriminates anomalies by measuring the deviations to the learned distribution. However, conventional representation learning is only a crude description of video normality and lacks an exploration of its underlying causality. The learned statistical dependence is unreliable for diverse regular events in the real world and may cause high false alarms due to overgeneralization. Inspired by causal representation learning, we think that there exists a causal variable capable of adequately representing the general patterns of regular events in which anomalies will present significant variations. Therefore, we design a causality-inspired representation consistency (CRC) framework to implicitly learn the unobservable causal variables of normality directly from available normal videos and detect abnormal events with the learned representation consistency. Extensive experiments show that the causality-inspired normality is robust to regular events with label-independent shifts, and the proposed CRC framework can quickly and accurately detect various complicated anomalies from real-world surveillance videos.

cs.MM↗
arXiv2023

AIDE: A Vision-Driven Multi-View, Multi-Modal, Multi-Tasking Dataset for Assistive Driving Perception

Driver distraction has become a significant cause of severe traffic accidents over the past decade. Despite the growing development of vision-driven driver monitoring systems, the lack of comprehensive perception datasets restricts road safety and traffic security. In this paper, we present an AssIstive Driving pErception dataset (AIDE) that considers context information both inside and outside the vehicle in naturalistic scenarios. AIDE facilitates holistic driver monitoring through three distinctive characteristics, including multi-view settings of driver and scene, multi-modal annotations of face, body, posture, and gesture, and four pragmatic task designs for driving understanding. To thoroughly explore AIDE, we provide experimental benchmarks on three kinds of baseline frameworks via extensive methods. Moreover, two fusion strategies are introduced to give new insights into learning effective multi-stream/modal representations. We also systematically investigate the importance and rationality of the key components in AIDE and benchmarks. The project link is https://github.com/ydk122024/AIDE.

cs.CV↗
arXiv2023

ARPES Detection of Superconducting Gap Sign in Unconventional Superconductors

Superconductivity is realized by opening a gap in the superconducting state. The gap symmetry is crucial in understanding the underlying superconductivity mechanism. The magnitude and the phase are essential in fully characterizing the superconducting gap. Angle-resolved photoemission spectroscopy (ARPES) has played a key role in determining the gap symmetry in unconventional superconductors. However, it has been considered so far that ARPES can only measure the magnitude of the superconducting gap but not its phase; the phase has to be detected by other phase-sensitive techniques. Here we propose a new method to directly detect the superconducting gap sign by using ARPES. This method is successfully validated in a cuprate superconductor with a well-known $d$-wave gap symmetry. When two bands are nearby in momentum space and have a strong interband interaction, the resulted electronic structures in the superconducting state are sensitive to the relative gap sign between the two bands which can be captured by ARPES measurements. Our present work provides a new way to detect the gap sign and can be applied to various superconductors, particularly those with multiple orbitals like the iron-based superconductors. It also makes ARPES more powerful to determine both the gap magnitude and the phase that are significant in understanding the superconductivity mechanism of unconventional superconductors.

cond-mat.supr-con↗
arXiv2023

B2Opt: Learning to Optimize Black-box Optimization with Little Budget

The core challenge of high-dimensional and expensive black-box optimization (BBO) is how to obtain better performance faster with little function evaluation cost. The essence of the problem is how to design an efficient optimization strategy tailored to the target task. This paper designs a powerful optimization framework to automatically learn the optimization strategies from the target or cheap surrogate task without human intervention. However, current methods are weak for this due to poor representation of optimization strategy. To achieve this, 1) drawing on the mechanism of genetic algorithm, we propose a deep neural network framework called B2Opt, which has a stronger representation of optimization strategies based on survival of the fittest; 2) B2Opt can utilize the cheap surrogate functions of the target task to guide the design of the efficient optimization strategies. Compared to the state-of-the-art BBO baselines, B2Opt can achieve multiple orders of magnitude performance improvement with less function evaluation cost. We validate our proposal on high-dimensional synthetic functions and two real-world applications. We also find that deep B2Opt performs better than shallow ones.

cs.LG↗
arXiv2023

Network Collaborator: Knowledge Transfer Between Network Reconstruction and Community Detection

This paper focuses on jointly inferring network and community structures from the dynamics of complex systems. Although many approaches have been designed to solve these two problems solely, none of them consider explicit shareable knowledge across these two tasks. Community detection (CD) from dynamics and network reconstruction (NR) from dynamics are natural synergistic tasks that motivate the proposed evolutionary multitasking NR and CD framework, called network collaborator (NC). In the process of NC, the NR task explicitly transfers several better network structures for the CD task, and the CD task explicitly transfers a better community structure to assist the NR task. Moreover, to transfer knowledge from the NR task to the CD task, NC models the study of CD from dynamics to find communities in the dynamic network and then considers whether to transfer knowledge across tasks. A test suite for multitasking NR and CD problems (MTNRCDPs) is designed to verify the performance of NC. The experimental results conducted on the designed MTNRCDPs have demonstrated that joint NR with CD has a synergistic effect, where the network structure used to inform the existence of communities is also inherently employed to improve the reconstruction accuracy, which, in turn, can better demonstrate the discovering of the community structure. The code is available at: https://github.com/xiaofangxd/EMTNRCD.

cs.SI↗
arXiv2023

Perceptual Quality Assessment of Omnidirectional Audio-visual Signals

Omnidirectional videos (ODVs) play an increasingly important role in the application fields of medical, education, advertising, tourism, etc. Assessing the quality of ODVs is significant for service-providers to improve the user's Quality of Experience (QoE). However, most existing quality assessment studies for ODVs only focus on the visual distortions of videos, while ignoring that the overall QoE also depends on the accompanying audio signals. In this paper, we first establish a large-scale audio-visual quality assessment dataset for omnidirectional videos, which includes 375 distorted omnidirectional audio-visual (A/V) sequences generated from 15 high-quality pristine omnidirectional A/V contents, and the corresponding perceptual audio-visual quality scores. Then, we design three baseline methods for full-reference omnidirectional audio-visual quality assessment (OAVQA), which combine existing state-of-the-art single-mode audio and video QA models via multimodal fusion strategies. We validate the effectiveness of the A/V multimodal fusion method for OAVQA on our dataset, which provides a new benchmark for omnidirectional QoE evaluation. Our dataset is available at https://github.com/iamazxl/OAVQA.

cs.CV↗
arXiv2023

Discover governing differential equations from evolving systems

Discovering the governing equations of evolving systems from available observations is essential and challenging. In this paper, we consider a new scenario: discovering governing equations from streaming data. Current methods struggle to discover governing differential equations with considering measurements as a whole, leading to failure to handle this task. We propose an online modeling method capable of handling samples one by one sequentially by modeling streaming data instead of processing the entire dataset. The proposed method performs well in discovering ordinary differential equations (ODEs) and partial differential equations (PDEs) from streaming data. Evolving systems are changing over time, which invariably changes with system status. Thus, finding the exact change points is critical. The measurement generated from a changed system is distributed dissimilarly to before; hence, the difference can be identified by the proposed method. Our proposal is competitive in identifying the change points and discovering governing differential equations in three hybrid systems and two switching linear systems.

physics.comp-ph↗
arXiv2023

AIGCIQA2023: A Large-scale Image Quality Assessment Database for AI Generated Images: from the Perspectives of Quality, Authenticity and Correspondence

In this paper, in order to get a better understanding of the human visual preferences for AIGIs, a large-scale IQA database for AIGC is established, which is named as AIGCIQA2023. We first generate over 2000 images based on 6 state-of-the-art text-to-image generation models using 100 prompts. Based on these images, a well-organized subjective experiment is conducted to assess the human visual preferences for each image from three perspectives including quality, authenticity and correspondence. Finally, based on this large-scale database, we conduct a benchmark experiment to evaluate the performance of several state-of-the-art IQA metrics on our constructed database.

cs.CV↗
arXiv2023

Quantum dynamic response-based NV-diamond magnetometry: Robustness to decoherence and applications in motion detection of magnetic nanoparticles

We propose a novel quantum sensing protocol that leverages the dynamical response of physical observables to quenches in quantum systems. Specifically, we use the nitrogen-vacancy (NV) color center in diamond to realize both scalar and vector magnetometry via quantum response. Furthermore, we suggest a method for detecting the motion of magnetic nanoparticles, which is challenging with conventional interference-based sensors. To achieve this, we derive the closed exact form of the Berry curvature corresponding to NV centers and design quenching protocols to extract the Berry curvature via dynamical response. By constructing and solving non-linear equations, the magnetic field and instantaneous motion velocity of the magnetic nanoparticle can be deduced. We investigate the feasibility of our sensing scheme in the presence of decoherence and show through numerical simulations that it is robust to decoherence. Intriguingly, we have observed that a vanishing nuclear spin polarization in diamond actually benefits our dynamic sensing scheme, which stands in contrast to conventional Ramsey-based schemes. In comparison to Ramsey-based sensing schemes, our proposed scheme can sense an arbitrary time-dependent magnetic field, as long as its time dependence is nearly adiabatic.

quant-ph↗
arXiv2023

Zn-doped P-type InAs Nanocrystal Quantum Dots

Doped heavy metal-free III-V semiconductor nanocrystal quantum dots are of great interest both from the fundamental aspects of doping in highly confined structures, and from the applicative side of utilizing such building blocks in the fabrication of p-n homojunction devices. InAs nanocrystals, that are of particular relevance for short wave IR detection and emission applications, manifest heavy n-type character poising a challenge for their transition to p-type behavior. We present p-type doping of InAs nanocrystals with Zn-enabling control over the charge carrier type in InAs QDs field effect transistors. The post-synthesis doping reaction mechanism is studied for Zn precursors with varying reactivity. Successful p-type doping was achieved by the more reactive precursor, diethylzinc. Substitutional doping by Zn2+ replacing In3+ is established by X-ray absorption spectroscopy analysis. Furthermore, enhanced near IR photoluminescence is observed due to surface passivation by Zn as indicated from elemental mapping utilizing high resolution electron microscopy corroborated by X-ray photoelectron spectroscopy study. The demonstrated ability to control the carrier type, along with the improved emission characteristics, paves the way towards fabrication of optoelectronic devices active in the short wave IR region utilizing heavy-metal free nanocrystal building blocks.

cond-mat.mes-hall↗
arXiv2023

Less Learn Shortcut: Analyzing and Mitigating Learning of Spurious Feature-Label Correlation

Recent research has revealed that deep neural networks often take dataset biases as a shortcut to make decisions rather than understand tasks, leading to failures in real-world applications. In this study, we focus on the spurious correlation between word features and labels that models learn from the biased data distribution of training data. In particular, we define the word highly co-occurring with a specific label as biased word, and the example containing biased word as biased example. Our analysis shows that biased examples are easier for models to learn, while at the time of prediction, biased words make a significantly higher contribution to the models' predictions, and models tend to assign predicted labels over-relying on the spurious correlation between words and labels. To mitigate models' over-reliance on the shortcut (i.e. spurious correlation), we propose a training strategy Less-Learn-Shortcut (LLS): our strategy quantifies the biased degree of the biased examples and down-weights them accordingly. Experimental results on Question Matching, Natural Language Inference and Sentiment Analysis tasks show that LLS is a task-agnostic strategy and can improve the model performance on adversarial data while maintaining good performance on in-domain data.

cs.CL↗
arXiv2023

Description-Enhanced Label Embedding Contrastive Learning for Text Classification

Text Classification is one of the fundamental tasks in natural language processing, which requires an agent to determine the most appropriate category for input sentences. Recently, deep neural networks have achieved impressive performance in this area, especially Pre-trained Language Models (PLMs). Usually, these methods concentrate on input sentences and corresponding semantic embedding generation. However, for another essential component: labels, most existing works either treat them as meaningless one-hot vectors or use vanilla embedding methods to learn label representations along with model training, underestimating the semantic information and guidance that these labels reveal. To alleviate this problem and better exploit label information, in this paper, we employ Self-Supervised Learning (SSL) in model learning process and design a novel self-supervised Relation of Relation (R2) classification task for label utilization from a one-hot manner perspective. Then, we propose a novel Relation of Relation Learning Network (R2-Net) for text classification, in which text classification and R2 classification are treated as optimization targets. Meanwhile, triplet loss is employed to enhance the analysis of differences and connections among labels. Moreover, considering that one-hot usage is still short of exploiting label information, we incorporate external knowledge from WordNet to obtain multi-aspect descriptions for label semantic learning and extend R2-Net to a novel Description-Enhanced Label Embedding network (DELE) from a label embedding perspective. ...

cs.CL↗
arXiv2023

COSA: Concatenated Sample Pretrained Vision-Language Foundation Model

Due to the limited scale and quality of video-text training corpus, most vision-language foundation models employ image-text datasets for pretraining and primarily focus on modeling visually semantic representations while disregarding temporal semantic representations and correlations. To address this issue, we propose COSA, a COncatenated SAmple pretrained vision-language foundation model. COSA jointly models visual contents and event-level temporal cues using only image-text corpora. We achieve this by sequentially concatenating multiple image-text pairs as inputs for pretraining. This transformation effectively converts existing image-text corpora into a pseudo long-form video-paragraph corpus, enabling richer scene transformations and explicit event-description correspondence. Extensive experiments demonstrate that COSA consistently improves performance across a broad range of downstream tasks, including long-form/short-form video-text tasks and image-text tasks such as retrieval, captioning, and question answering. Notably, COSA achieves state-of-the-art results on various competitive benchmarks. Code and model are released at https://github.com/TXH-mercury/COSA.

cs.CV↗
arXiv2023

MAMO: Masked Multimodal Modeling for Fine-Grained Vision-Language Representation Learning

Multimodal representation learning has shown promising improvements on various vision-language tasks. Most existing methods excel at building global-level alignment between vision and language while lacking effective fine-grained image-text interaction. In this paper, we propose a jointly masked multimodal modeling method to learn fine-grained multimodal representations. Our method performs joint masking on image-text input and integrates both implicit and explicit targets for the masked signals to recover. The implicit target provides a unified and debiased objective for vision and language, where the model predicts latent multimodal representations of the unmasked input. The explicit target further enriches the multimodal representations by recovering high-level and semantically meaningful information: momentum visual features of image patches and concepts of word tokens. Through such a masked modeling process, our model not only learns fine-grained multimodal interaction, but also avoids the semantic gap between high-level representations and low- or mid-level prediction targets (e.g. image pixels), thus producing semantically rich multimodal representations that perform well on both zero-shot and fine-tuned settings. Our pre-trained model (named MAMO) achieves state-of-the-art performance on various downstream vision-language tasks, including image-text retrieval, visual question answering, visual reasoning, and weakly-supervised visual grounding.

cs.CV↗
arXiv2023

Variation of the Atmospheric Boundary Layer Height at the Eastern Edge of the Tibetan Plateau

This paper utilized the high temporal and spatial resolution temperature profile data observed by the multi-channel microwave radiometer at the Large High Altitude Air Shower Observatory (LHAASO) on the eastern slope of the Tibetan Plateau from February to May and August to November 2021, combined with the ERA5 reanalysis data products for the whole year of 2021, to study the daily, monthly, and seasonal variations of the atmospheric boundary layer height (ABLH). The results are as follows: (1) The ABLH on sunny days showed obvious fluctuations with peaks and valleys. The ABLH continued to rise with the increase of surface temperature after sunrise and usually reached its maximum value in the afternoon around 18:00, then rapidly decreased until sunset. (2) The average ABLH in April was the highest at about 1200 m, while it was only around 600 m in November. The ABLH fluctuated greatly during the day and was stable at around 400 m at night. The ABLH results obtained from ERA5 were slightly smaller overall but had a consistent trend of change with the microwave radiometer. (3) The maximum ABLH appeared in spring, followed by summer and autumn, and winter had the lowest value, with all peaks reached around 14:00-15:00. These results are of great significance for understanding the ABLH on the eastern slope of the Tibetan Plateau, and provide reference for the absolute calibration of photon numbers of the LHAASO telescope and the atmospheric monitoring plan, as well as for evaluating the authenticity and accuracy of existing reanalysis datasets.

physics.ao-ph↗
arXiv2023

ChatBridge: Bridging Modalities with Large Language Model as a Language Catalyst

Building general-purpose models that can perceive diverse real-world modalities and solve various tasks is an appealing target in artificial intelligence. In this paper, we present ChatBridge, a novel multimodal language model that leverages the expressive capabilities of language as the catalyst to bridge the gap between various modalities. We show that only language-paired two-modality data is sufficient to connect all modalities. ChatBridge leverages recent large language models (LLM) and extends their zero-shot capabilities to incorporate diverse multimodal inputs. ChatBridge undergoes a two-stage training. The first stage aligns each modality with language, which brings emergent multimodal correlation and collaboration abilities. The second stage instruction-finetunes ChatBridge to align it with user intent with our newly proposed multimodal instruction tuning dataset, named MULTIS, which covers a wide range of 16 multimodal tasks of text, image, video, and audio modalities. We show strong quantitative and qualitative results on zero-shot multimodal tasks covering text, image, video, and audio modalities. All codes, data, and models of ChatBridge will be open-sourced.

cs.CV↗
arXiv2023

MMNet: Multi-Mask Network for Referring Image Segmentation

Referring image segmentation aims to segment an object referred to by natural language expression from an image. However, this task is challenging due to the distinct data properties between text and image, and the randomness introduced by diverse objects and unrestricted language expression. Most of previous work focus on improving cross-modal feature fusion while not fully addressing the inherent uncertainty caused by diverse objects and unrestricted language. To tackle these problems, we propose an end-to-end Multi-Mask Network for referring image segmentation(MMNet). we first combine picture and language and then employ an attention mechanism to generate multiple queries that represent different aspects of the language expression. We then utilize these queries to produce a series of corresponding segmentation masks, assigning a score to each mask that reflects its importance. The final result is obtained through the weighted sum of all masks, which greatly reduces the randomness of the language expression. Our proposed framework demonstrates superior performance compared to state-of-the-art approaches on the two most commonly used datasets, RefCOCO, RefCOCO+ and G-Ref, without the need for any post-processing. This further validates the efficacy of our proposed framework.

cs.CV↗
arXiv2023

VLAB: Enhancing Video Language Pre-training by Feature Adapting and Blending

Large-scale image-text contrastive pre-training models, such as CLIP, have been demonstrated to effectively learn high-quality multimodal representations. However, there is limited research on learning video-text representations for general video multimodal tasks based on these powerful features. Towards this goal, we propose a novel video-text pre-training method dubbed VLAB: Video Language pre-training by feature Adapting and Blending, which transfers CLIP representations to video pre-training tasks and develops unified video multimodal models for a wide range of video-text tasks. Specifically, VLAB is founded on two key strategies: feature adapting and feature blending. In the former, we introduce a new video adapter module to address CLIP's deficiency in modeling temporal information and extend the model's capability to encompass both contrastive and generative tasks. In the latter, we propose an end-to-end training method that further enhances the model's performance by exploiting the complementarity of image and video features. We validate the effectiveness and versatility of VLAB through extensive experiments on highly competitive video multimodal tasks, including video text retrieval, video captioning, and video question answering. Remarkably, VLAB outperforms competing methods significantly and sets new records in video question answering on MSRVTT, MSVD, and TGIF datasets. It achieves an accuracy of 49.6, 61.0, and 79.0, respectively. Codes and models will be released.

cs.CV↗