arXiv subjects

Yuhan Zhang

Publications and source records attributed to Yuhan Zhang.

69 records · Page 4Linked to original sources

arXiv2023

Acoustic Scene Clustering Using Joint Optimization of Deep Embedding Learning and Clustering Iteration

Recent efforts have been made on acoustic scene classification in the audio signal processing community. In contrast, few studies have been conducted on acoustic scene clustering, which is a newly emerging problem. Acoustic scene clustering aims at merging the audio recordings of the same class of acoustic scene into a single cluster without using prior information and training classifiers. In this study, we propose a method for acoustic scene clustering that jointly optimizes the procedures of feature learning and clustering iteration. In the proposed method, the learned feature is a deep embedding that is extracted from a deep convolutional neural network (CNN), while the clustering algorithm is the agglomerative hierarchical clustering (AHC). We formulate a unified loss function for integrating and optimizing these two procedures. Various features and methods are compared. The experimental results demonstrate that the proposed method outperforms other unsupervised methods in terms of the normalized mutual information and the clustering accuracy. In addition, the deep embedding outperforms many state-of-the-art features.

eess.AS↗

arXiv2023

Self-supervised arbitrary scale super-resolution framework for anisotropic MRI

In this paper, we propose an efficient self-supervised arbitrary-scale super-resolution (SR) framework to reconstruct isotropic magnetic resonance (MR) images from anisotropic MRI inputs without involving external training data. The proposed framework builds a training dataset using in-the-wild anisotropic MR volumes with arbitrary image resolution. We then formulate the 3D volume SR task as a SR problem for 2D image slices. The anisotropic volume's high-resolution (HR) plane is used to build the HR-LR image pairs for model training. We further adapt the implicit neural representation (INR) network to implement the 2D arbitrary-scale image SR model. Finally, we leverage the well-trained proposed model to up-sample the 2D LR plane extracted from the anisotropic MR volumes to their HR views. The isotropic MR volumes thus can be reconstructed by stacking and averaging the generated HR slices. Our proposed framework has two major advantages: (1) It only involves the arbitrary-resolution anisotropic MR volumes, which greatly improves the model practicality in real MR imaging scenarios (e.g., clinical brain image acquisition); (2) The INR-based SR model enables arbitrary-scale image SR from the arbitrary-resolution input image, which significantly improves model training efficiency. We perform experiments on a simulated public adult brain dataset and a real collected 7T brain dataset. The results indicate that our current framework greatly outperforms two well-known self-supervised models for anisotropic MR image SR tasks.

eess.IV↗

arXiv2023

Adaptive Data Augmentation for Contrastive Learning

In computer vision, contrastive learning is the most advanced unsupervised learning framework. Yet most previous methods simply apply fixed composition of data augmentations to improve data efficiency, which ignores the changes in their optimal settings over training. Thus, the pre-determined parameters of augmentation operations cannot always fit well with an evolving network during the whole training period, which degrades the quality of the learned representations. In this work, we propose AdDA, which implements a closed-loop feedback structure to a generic contrastive learning network. AdDA works by allowing the network to adaptively adjust the augmentation compositions according to the real-time feedback. This online adjustment helps maintain the dynamic optimal composition and enables the network to acquire more generalizable representations with minimal computational overhead. AdDA achieves competitive results under the common linear protocol on ImageNet-100 classification (+1.11% on MoCo v2).

cs.CV↗

arXiv2023

Feedback Effect in User Interaction with Intelligent Assistants: Delayed Engagement, Adaption and Drop-out

With the growing popularity of intelligent assistants (IAs), evaluating IA quality becomes an increasingly active field of research. This paper identifies and quantifies the feedback effect, a novel component in IA-user interactions: how the capabilities and limitations of the IA influence user behavior over time. First, we demonstrate that unhelpful responses from the IA cause users to delay or reduce subsequent interactions in the short term via an observational study. Next, we expand the time horizon to examine behavior changes and show that as users discover the limitations of the IA's understanding and functional capabilities, they learn to adjust the scope and wording of their requests to increase the likelihood of receiving a helpful response from the IA. Our findings highlight the impact of the feedback effect at both the micro and meso levels. We further discuss its macro-level consequences: unsatisfactory interactions continuously reduce the likelihood and diversity of future user engagements in a feedback loop.

cs.HC↗

arXiv2023

Semantic Interleaving Global Channel Attention for Multilabel Remote Sensing Image Classification

Multi-Label Remote Sensing Image Classification (MLRSIC) has received increasing research interest. Taking the cooccurrence relationship of multiple labels as additional information helps to improve the performance of this task. Current methods focus on using it to constrain the final feature output of a Convolutional Neural Network (CNN). On the one hand, these methods do not make full use of label correlation to form feature representation. On the other hand, they increase the label noise sensitivity of the system, resulting in poor robustness. In this paper, a novel method called Semantic Interleaving Global Channel Attention (SIGNA) is proposed for MLRSIC. First, the label co-occurrence graph is obtained according to the statistical information of the data set. The label co-occurrence graph is used as the input of the Graph Neural Network (GNN) to generate optimal feature representations. Then, the semantic features and visual features are interleaved, to guide the feature expression of the image from the original feature space to the semantic feature space with embedded label relations. SIGNA triggers global attention of feature maps channels in a new semantic feature space to extract more important visual features. Multihead SIGNA based feature adaptive weighting networks are proposed to act on any layer of CNN in a plug-and-play manner. For remote sensing images, better classification performance can be achieved by inserting CNN into the shallow layer. We conduct extensive experimental comparisons on three data sets: UCM data set, AID data set, and DFC15 data set. Experimental results demonstrate that the proposed SIGNA achieves superior classification performance compared to state-of-the-art (SOTA) methods. It is worth mentioning that the codes of this paper will be open to the community for reproducibility research. Our codes are available at https://github.com/kyle-one/SIGNA.

cs.CV↗

arXiv2022

OCTA-500: A Retinal Dataset for Optical Coherence Tomography Angiography Study

Optical coherence tomography angiography (OCTA) is a novel imaging modality that has been widely utilized in ophthalmology and neuroscience studies to observe retinal vessels and microvascular systems. However, publicly available OCTA datasets remain scarce. In this paper, we introduce the largest and most comprehensive OCTA dataset dubbed OCTA-500, which contains OCTA imaging under two fields of view (FOVs) from 500 subjects. The dataset provides rich images and annotations including two modalities (OCT/OCTA volumes), six types of projections, four types of text labels (age / gender / eye / disease) and seven types of segmentation labels (large vessel/capillary/artery/vein/2D FAZ/3D FAZ/retinal layers). Then, we propose a multi-object segmentation task called CAVF, which integrates capillary segmentation, artery segmentation, vein segmentation, and FAZ segmentation under a unified framework. In addition, we optimize the 3D-to-2D image projection network (IPN) to IPN-V2 to serve as one of the segmentation baselines. Experimental results demonstrate that IPN-V2 achieves an ~10% mIoU improvement over IPN on CAVF task. Finally, we further study the impact of several dataset characteristics: the training set size, the model input (OCT/OCTA, 3D volume/2D projection), the baseline networks, and the diseases. The dataset and code are publicly available at: https://ieee-dataport.org/open-access/octa-500.

eess.IV↗

arXiv2022

Representing Affect Information in Word Embeddings

A growing body of research in natural language processing (NLP) and natural language understanding (NLU) is investigating human-like knowledge learned or encoded in the word embeddings from large language models. This is a step towards understanding what knowledge language models capture that resembles human understanding of language and communication. Here, we investigated whether and how the affect meaning of a word (i.e., valence, arousal, dominance) is encoded in word embeddings pre-trained in large neural networks. We used the human-labeled dataset as the ground truth and performed various correlational and classification tests on four types of word embeddings. The embeddings varied in being static or contextualized, and how much affect specific information was prioritized during the pre-training and fine-tuning phase. Our analyses show that word embedding from the vanilla BERT model did not saliently encode the affect information of English words. Only when the BERT model was fine-tuned on emotion-related tasks or contained extra contextualized information from emotion-rich contexts could the corresponding embedding encode more relevant affect information.

cs.CL↗

arXiv2021

Dynamic Metric Learning: Towards a Scalable Metric Space to Accommodate Multiple Semantic Scales

This paper introduces a new fundamental characteristic, \ie, the dynamic range, from real-world metric tools to deep visual recognition. In metrology, the dynamic range is a basic quality of a metric tool, indicating its flexibility to accommodate various scales. Larger dynamic range offers higher flexibility. In visual recognition, the multiple scale problem also exist. Different visual concepts may have different semantic scales. For example, ``Animal'' and ``Plants'' have a large semantic scale while ``Elk'' has a much smaller one. Under a small semantic scale, two different elks may look quite \emph{different} to each other . However, under a large semantic scale (\eg, animals and plants), these two elks should be measured as being \emph{similar}. %We argue that such flexibility is also important for deep metric learning, because different visual concepts indeed correspond to different semantic scales. Introducing the dynamic range to deep metric learning, we get a novel computer vision task, \ie, the Dynamic Metric Learning. It aims to learn a scalable metric space to accommodate visual concepts across multiple semantic scales. Based on three types of images, \emph{i.e.}, vehicle, animal and online products, we construct three datasets for Dynamic Metric Learning. We benchmark these datasets with popular deep metric learning methods and find Dynamic Metric Learning to be very challenging. The major difficulty lies in a conflict between different scales: the discriminative ability under a small scale usually compromises the discriminative ability under a large one, and vice versa. As a minor contribution, we propose Cross-Scale Learning (CSL) to alleviate such conflict. We show that CSL consistently improves the baseline on all the three datasets. The datasets and the code will be publicly available at https://github.com/SupetZYK/DynamicMetricLearning.

cs.CV↗

arXiv2020

Meta-Learning for Neural Relation Classification with Distant Supervision

Distant supervision provides a means to create a large number of weakly labeled data at low cost for relation classification. However, the resulting labeled instances are very noisy, containing data with wrong labels. Many approaches have been proposed to select a subset of reliable instances for neural model training, but they still suffer from noisy labeling problem or underutilization of the weakly-labeled data. To better select more reliable training instances, we introduce a small amount of manually labeled data as reference to guide the selection process. In this paper, we propose a meta-learning based approach, which learns to reweight noisy training data under the guidance of reference data. As the clean reference data is usually very small, we propose to augment it by dynamically distilling the most reliable elite instances from the noisy data. Experiments on several datasets demonstrate that the reference data can effectively guide the selection of training data, and our augmented approach consistently improves the performance of relation classification comparing to the existing state-of-the-art methods.

cs.CL↗

arXiv2020

Modeling the US-China trade conflict: a utility theory approach

This paper models the US-China trade conflict and attempts to analyze the (optimal) strategic choices. In contrast to the existing literature on the topic, we employ the expected utility theory and examine the conflict mathematically. In both perfect information and incomplete information games, we show that expected net gains diminish as the utility of winning increases because of the costs incurred during the struggle. We find that the best response function exists for China but not for the US during the conflict. We argue that the less the US coerces China to change its existing trade practices, the higher the US expected net gains. China's best choice is to maintain the status quo, and any further aggression in its policy and behavior will aggravate the situation.

econ.GN↗

arXiv2020

Circle Loss: A Unified Perspective of Pair Similarity Optimization

This paper provides a pair similarity optimization viewpoint on deep feature learning, aiming to maximize the within-class similarity $s_p$ and minimize the between-class similarity $s_n$. We find a majority of loss functions, including the triplet loss and the softmax plus cross-entropy loss, embed $s_n$ and $s_p$ into similarity pairs and seek to reduce $(s_n-s_p)$. Such an optimization manner is inflexible, because the penalty strength on every single similarity score is restricted to be equal. Our intuition is that if a similarity score deviates far from the optimum, it should be emphasized. To this end, we simply re-weight each similarity to highlight the less-optimized similarity scores. It results in a Circle loss, which is named due to its circular decision boundary. The Circle loss has a unified formula for two elemental deep feature learning approaches, i.e. learning with class-level labels and pair-wise labels. Analytically, we show that the Circle loss offers a more flexible optimization approach towards a more definite convergence target, compared with the loss functions optimizing $(s_n-s_p)$. Experimentally, we demonstrate the superiority of the Circle loss on a variety of deep feature learning tasks. On face recognition, person re-identification, as well as several fine-grained image retrieval datasets, the achieved performance is on par with the state of the art.

cs.CV↗

arXiv2016

Probing Carrier Transport and Structure-property Relationship of Highly Ordered Organic Semiconductors at Two-dimensional Limit

One of the basic assumptions in organic field-effect transistors, the most fundamental device unit in organic electronics, is that charge transport occurs two-dimensionally in the first few molecular layers near the dielectric interface. Although the mobility of bulk organic semiconductors has increased dramatically, direct probing of intrinsic charge transport in the two-dimensional limit has not been possible due to excessive disorders and traps in ultrathin organic thin films. Here, highly ordered mono- to tetra-layer pentacene crystals are realized by van der Waals (vdW) epitaxy on hexagonal BN. We find that the charge transport is dominated by hopping in the first conductive layer, but transforms to band-like in subsequent layers. Such abrupt phase transition is attributed to strong modulation of the molecular packing by interfacial vdW interactions, as corroborated by quantitative structural characterization and density functional theory calculations. The structural modulation becomes negligible beyond the second conductive layer, leading to a mobility saturation thickness of only ~3nm. Highly ordered organic ultrathin films provide a platform for new physics and device structures (such as heterostructures and quantum wells) that are not possible in conventional bulk crystals.

cond-mat.mes-hall↗

arXiv2015

A van der Waals pn heterojunction with organic/inorganic semiconductors

van der Waals (vdW) heterojunctions formed by two-dimensional (2D) materials have attracted tremendous attention due to their excellent electrical/optical properties and device applications. However, current 2D heterojunctions are largely limited to atomic crystals, and hybrid organic/inorganic structures are rarely explored. Here, we fabricate hybrid 2D heterostructures with p-type dioctylbenzothienobenzothiophene (C8-BTBT) and n-type MoS2. We find that few-layer C8-BTBT molecular crystals can be grown on monolayer MoS2 by vdW epitaxy, with pristine interface and controllable thickness down to monolayer. The operation of the C8-BTBT/MoS2 vertical heterojunction devices is highly tunable by bias and gate voltages between three different regimes: interfacial recombination, tunneling and blocking. The pn junction shows diode-like behavior with rectifying ratio up to 105 at the room temperature. Our devices also exhibit photovoltaic responses with power conversion efficiency of 0.31% and photoresponsivity of 22mA/W. With wide material combinations, such hybrid 2D structures will offer possibilities for opto-electronic devices that are not possible from individual constituents.

cond-mat.mes-hall↗

arXiv2014

Unveiling the Structural Origin of the High Carrier Mobility of a Molecular Monolayer on Boron Nitride

Very recently, it was demonstrated that the carrier mobility of a molecular monolayer dioctylbenzothienobenzothiophene (C8-BTBT) on boron nitride can reach 10 cm2/Vs, the highest among the previously reported monolayer molecular field-effect transistors. Here we show that the high-quality single crystal of the C8-BTBT monolayer may be the key origin of the record-high carrier mobility. We discover that the C8-BTBT molecules prefer layer-by-layer growth on both hexagonal boron nitride and graphene. The flatness of these substrates substantially decreases the C8-BTBT nucleation density and enables repeatable growth of large-area single crystal of the C8-BTBT monolayer. Our experimental result indicates that only out-of-plane roughness greater than 0.6 nm of the substrates could induce disturbance in the crystal growth and consequently affect the charge transport. This information would be important in guiding the growth of high-quality epitaxy molecular film.

cond-mat.mtrl-sci↗

arXiv2014

Two-dimensional Quasi-Freestanding Molecular Crystals for High-Performance Organic Field-Effect Transistors

Two-dimensional atomic crystals are extensively studied in recent years due to their exciting physics and device applications. However, a molecular counterpart, with scalable processability and competitive device performance, is still challenging. Here, we demonstrate that high-quality few-layer dioctylbenzothienobenzothiophene molecular crystals can be grown on graphene or boron nitride substrate via van der Waals epitaxy, with precisely controlled thickness down to monolayer, large-area single crystal, low process temperature and patterning capability. The crystalline layers are atomically smooth and effectively decoupled from the substrate due to weak van der Waals interactions, affording a pristine interface for high-performance organic transistors. As a result, monolayer dioctylbenzothienobenzothiophene molecular crystal field-effect transistors on boron nitride show record-high carrier mobility up to 10cm2V-1s-1 and aggressively scaled saturation voltage around 1V. Our work unveils an exciting new class of two-dimensional molecular materials for electronic and optoelectronic applications.

cond-mat.mtrl-sci↗