arXiv subjects

Li Zhang

Publications and source records attributed to Li Zhang.

At least 541 records · Page 30Linked to original sources

arXiv2023

CLIN: A Continually Learning Language Agent for Rapid Task Adaptation and Generalization

Language agents have shown some ability to interact with an external environment, e.g., a virtual world such as ScienceWorld, to perform complex tasks, e.g., growing a plant, without the startup costs of reinforcement learning. However, despite their zero-shot capabilities, these agents to date do not continually improve over time beyond performance refinement on a specific task. Here we present CLIN, the first language-based agent to achieve this, so that it continually improves over multiple trials, including when both the environment and task are varied, and without requiring parameter updates. Our approach is to use a persistent, dynamic, textual memory centered on causal abstractions (rather than general "helpful hints") that is regularly updated after each trial so that the agent gradually learns useful knowledge for new trials. In the ScienceWorld benchmark, CLIN is able to continually improve on repeated trials on the same task and environment, outperforming state-of-the-art reflective language agents like Reflexion by 23 absolute points. CLIN can also transfer its learning to new environments (or new tasks), improving its zero-shot performance by 4 points (13 for new tasks) and can further improve performance there through continual memory updates, enhancing performance by an additional 17 points (7 for new tasks). This suggests a new architecture for agents built on frozen models that can still continually and rapidly improve over time.

cs.CL↗

arXiv2023

CBARF: Cascaded Bundle-Adjusting Neural Radiance Fields from Imperfect Camera Poses

Existing volumetric neural rendering techniques, such as Neural Radiance Fields (NeRF), face limitations in synthesizing high-quality novel views when the camera poses of input images are imperfect. To address this issue, we propose a novel 3D reconstruction framework that enables simultaneous optimization of camera poses, dubbed CBARF (Cascaded Bundle-Adjusting NeRF).In a nutshell, our framework optimizes camera poses in a coarse-to-fine manner and then reconstructs scenes based on the rectified poses. It is observed that the initialization of camera poses has a significant impact on the performance of bundle-adjustment (BA). Therefore, we cascade multiple BA modules at different scales to progressively improve the camera poses. Meanwhile, we develop a neighbor-replacement strategy to further optimize the results of BA in each stage. In this step, we introduce a novel criterion to effectively identify poorly estimated camera poses. Then we replace them with the poses of neighboring cameras, thus further eliminating the impact of inaccurate camera poses. Once camera poses have been optimized, we employ a density voxel grid to generate high-quality 3D reconstructed scenes and images in novel views. Experimental results demonstrate that our CBARF model achieves state-of-the-art performance in both pose optimization and novel view synthesis, especially in the existence of large camera pose noise.

cs.CV↗

arXiv2023

Faithful Chain-of-Thought Reasoning

While Chain-of-Thought (CoT) prompting boosts Language Models' (LM) performance on a gamut of complex reasoning tasks, the generated reasoning chain does not necessarily reflect how the model arrives at the answer (aka. faithfulness). We propose Faithful CoT, a reasoning framework involving two stages: Translation (Natural Language query $\rightarrow$ symbolic reasoning chain) and Problem Solving (reasoning chain $\rightarrow$ answer), using an LM and a deterministic solver respectively. This guarantees that the reasoning chain provides a faithful explanation of the final answer. Aside from interpretability, Faithful CoT also improves empirical performance: it outperforms standard CoT on 9 of 10 benchmarks from 4 diverse domains, with a relative accuracy gain of 6.3% on Math Word Problems (MWP), 3.4% on Planning, 5.5% on Multi-hop Question Answering (QA), and 21.4% on Relational Inference. Furthermore, with GPT-4 and Codex, it sets the new state-of-the-art few-shot performance on 7 datasets (with 95.0+ accuracy on 6 of them), showing a strong synergy between faithfulness and accuracy.

cs.CL↗

arXiv2023

Private Matrix Factorization with Public Item Features

We consider the problem of training private recommendation models with access to public item features. Training with Differential Privacy (DP) offers strong privacy guarantees, at the expense of loss in recommendation quality. We show that incorporating public item features during training can help mitigate this loss in quality. We propose a general approach based on collective matrix factorization (CMF), that works by simultaneously factorizing two matrices: the user feedback matrix (representing sensitive data) and an item feature matrix that encodes publicly available (non-sensitive) item information. The method is conceptually simple, easy to tune, and highly scalable. It can be applied to different types of public item data, including: (1) categorical item features; (2) item-item similarities learned from public sources; and (3) publicly available user feedback. Furthermore, these data modalities can be collectively utilized to fully leverage public data. Evaluating our method on a standard DP recommendation benchmark, we find that using public item features significantly narrows the quality gap between private models and their non-private counterparts. As privacy constraints become more stringent, models rely more heavily on public side features for recommendation. This results in a smooth transition from collaborative filtering to item-based contextual recommendations.

cs.IR↗

arXiv2023

Semidiscrete optical vortex droplets in quasi-phase-matched photonic crystals

A new scheme for producing semidiscrete self-trapped vortices (\textquotedblleft swirling photon droplets\textquotedblright ) in photonic crystals with competing quadratic ($χ^{(2)}$) and self-defocusing cubic ($χ^{(3)}$) nonlinearities is proposed. The photonic crystal is designed with a striped structure, in the form of spatially periodic modulation of the $χ^{(2)}$ susceptibility, which is imposed by the quasi-phase-matching technique. Unlike previous realizations of semidiscrete optical modes in composite media, built as combinations of continuous and arrayed discrete waveguides, the semidiscrete vortex droplets are produced here in the fully continuous medium. This work reveals that the system supports two types of semidiscrete vortex droplets, \textit{viz}., onsite- and intersite-centered ones, which feature, respectively, odd and even numbers of stripes, $\mathcal{N}$. Stability areas for the states with different values of $\mathcal{N}$ are identified in the system's parameter space. Some stability areas overlap with each others, giving rise to multistability of states with different $\mathcal{N}$. The coexisting states are mutually degenerate, featuring equal values of the Hamiltonian and propagation constant. An experimental scheme to realize the droplets is outlined, suggesting new possibilities for the long-distance transmission of structured light carrying orbital angular momentum in nonlinear media.

physics.optics↗

arXiv2023

Improved Prognostic Prediction of Pancreatic Cancer Using Multi-Phase CT by Integrating Neural Distance and Texture-Aware Transformer

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal cancer in which the tumor-vascular involvement greatly affects the resectability and, thus, overall survival of patients. However, current prognostic prediction methods fail to explicitly and accurately investigate relationships between the tumor and nearby important vessels. This paper proposes a novel learnable neural distance that describes the precise relationship between the tumor and vessels in CT images of different patients, adopting it as a major feature for prognosis prediction. Besides, different from existing models that used CNNs or LSTMs to exploit tumor enhancement patterns on dynamic contrast-enhanced CT imaging, we improved the extraction of dynamic tumor-related texture features in multi-phase contrast-enhanced CT by fusing local and global features using CNN and transformer modules, further enhancing the features extracted across multi-phase CT images. We extensively evaluated and compared the proposed method with existing methods in the multi-center (n=4) dataset with 1,070 patients with PDAC, and statistical analysis confirmed its clinical effectiveness in the external test set consisting of three centers. The developed risk marker was the strongest predictor of overall survival among preoperative factors and it has the potential to be combined with established clinical factors to select patients at higher risk who might benefit from neoadjuvant therapy.

eess.IV↗

arXiv2023

Designs and Implementations in Neural Network-based Video Coding

The past decade has witnessed the huge success of deep learning in well-known artificial intelligence applications such as face recognition, autonomous driving, and large language model like ChatGPT. Recently, the application of deep learning has been extended to a much wider range, with neural network-based video coding being one of them. Neural network-based video coding can be performed at two different levels: embedding neural network-based (NN-based) coding tools into a classical video compression framework or building the entire compression framework upon neural networks. This paper elaborates some of the recent exploration efforts of JVET (Joint Video Experts Team of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC29) in the name of neural network-based video coding (NNVC), falling in the former category. Specifically, this paper discusses two major NN-based video coding technologies, i.e. neural network-based intra prediction and neural network-based in-loop filtering, which have been investigated for several meeting cycles in JVET and finally adopted into the reference software of NNVC. Extensive experiments on top of the NNVC have been conducted to evaluate the effectiveness of the proposed techniques. Compared with VTM-11.0_nnvc, the proposed NN-based coding tools in NNVC-4.0 could achieve {11.94%, 21.86%, 22.59%}, {9.18%, 19.76%, 20.92%}, and {10.63%, 21.56%, 23.02%} BD-rate reductions on average for {Y, Cb, Cr} under random-access, low-delay, and all-intra configurations respectively.

eess.IV↗

arXiv2023

LB-SimTSC: An Efficient Similarity-Aware Graph Neural Network for Semi-Supervised Time Series Classification

Time series classification is an important data mining task that has received a lot of interest in the past two decades. Due to the label scarcity in practice, semi-supervised time series classification with only a few labeled samples has become popular. Recently, Similarity-aware Time Series Classification (SimTSC) is proposed to address this problem by using a graph neural network classification model on the graph generated from pairwise Dynamic Time Warping (DTW) distance of batch data. It shows excellent accuracy and outperforms state-of-the-art deep learning models in several few-label settings. However, since SimTSC relies on pairwise DTW distances, the quadratic complexity of DTW limits its usability to only reasonably sized datasets. To address this challenge, we propose a new efficient semi-supervised time series classification technique, LB-SimTSC, with a new graph construction module. Instead of using DTW, we propose to utilize a lower bound of DTW, LB_Keogh, to approximate the dissimilarity between instances in linear time, while retaining the relative proximity relationships one would have obtained via computing DTW. We construct the pairwise distance matrix using LB_Keogh and build a graph for the graph neural network. We apply this approach to the ten largest datasets from the well-known UCR time series classification archive. The results demonstrate that this approach can be up to 104x faster than SimTSC when constructing the graph on large datasets without significantly decreasing classification accuracy.

cs.LG↗

arXiv2023

Timbre-reserved Adversarial Attack in Speaker Identification

As a type of biometric identification, a speaker identification (SID) system is confronted with various kinds of attacks. The spoofing attacks typically imitate the timbre of the target speakers, while the adversarial attacks confuse the SID system by adding a well-designed adversarial perturbation to an arbitrary speech. Although the spoofing attack copies a similar timbre as the victim, it does not exploit the vulnerability of the SID model and may not make the SID system give the attacker's desired decision. As for the adversarial attack, despite the SID system can be led to a designated decision, it cannot meet the specified text or speaker timbre requirements for the specific attack scenarios. In this study, to make the attack in SID not only leverage the vulnerability of the SID model but also reserve the timbre of the target speaker, we propose a timbre-reserved adversarial attack in the speaker identification. We generate the timbre-reserved adversarial audios by adding an adversarial constraint during the different training stages of the voice conversion (VC) model. Specifically, the adversarial constraint is using the target speaker label to optimize the adversarial perturbation added to the VC model representations and is implemented by a speaker classifier joining in the VC model training. The adversarial constraint can help to control the VC model to generate the speaker-wised audio. Eventually, the inference of the VC model is the ideal adversarial fake audio, which is timbre-reserved and can fool the SID system.

cs.SD↗

arXiv2023

Nanopore-patterned CuSe drives the realization of PbSe-CuSe lateral heterostructure

Monolayer PbSe has been predicted to be a two-dimensional (2D) topological crystalline insulator (TCI) with crystalline symmetry-protected Dirac-cone-like edge states. Recently, few-layered epitaxial PbSe has been grown on the SrTiO3 substrate successfully, but the corresponding signature of the TCI was only observed for films not thinner than seven monolayers, largely due to interfacial strain. Here, we demonstrate a two-step method based on molecular beam epitaxy for the growth of the PbSe-CuSe lateral heterostructure on the Cu(111) substrate, in which we observe a nanopore patterned CuSe layer that acts as the template for lateral epitaxial growth of PbSe. This further results in a monolayer PbSe-CuSe lateral heterostructure with an atomically sharp interface. Scanning tunneling microscopy and spectroscopy measurements reveal a four-fold symmetric square lattice of such monolayer PbSe with a quasi-particle band gap of 1.8 eV, a value highly comparable with the theoretical value of freestanding PbSe. The weak monolayer-substrate interaction is further supported by both density functional theory (DFT) and projected crystal orbital Hamilton population, with the former predicting the monolayer's anti-bond state to reside below the Fermi level. Our work demonstrates a practical strategy to fabricate a high-quality in-plane heterostructure, involving a monolayer TCI, which is viable for further exploration of the topology-derived quantum physics and phenomena in the monolayer limit.

cond-mat.mtrl-sci↗

arXiv2023

Two-dimensional germanium islands with Dirac signature on Ag2Ge surface alloy

Two-dimensional (2D) Dirac materials have attracted intense research efforts due to their promise for applications ranging from field-effect transistors and low-power electronics to fault-tolerant quantum computation. One key challenge is to fabricate 2D Dirac materials hosting Dirac electrons. Here, monolayer germanene is successfully fabricated on a Ag2Ge surface alloy. Scanning tunneling spectroscopy measurements revealed a linear energy dispersion relation. The latter was supported by density functional theory calculations. These results demonstrate that monolayer germanene can be realistically fabricated on a Ag2Ge surface alloy. The finding opens the door to exploration and study of 2D Dirac material physics and device applications.

cond-mat.mtrl-sci↗

arXiv2023

Native Pb vacancy defects induced p-type characteristic in epitaxial monolayer PbSe

PbSe, a predicted two-dimensional (2D) topological crystalline insulator (TCI) in the monolayer limit, possess excellent thermoelectric and infrared optical properties. Native defects in PbSe take a crucial role for the applications. However, little attention has been paid to the defect induced doping characteristics. Here, we provide an experimental and theoretical investigation of defects induced p-type characteristic on epitaxial monolayer PbSe on Au(111). Scanning tunneling microscopy (STM) measurements demonstrate an epitaxial PbSe monolayer with a fourfold symmetric lattice. Combined scanning tunneling spectroscopy (STS) and density functional theory (DFT) calculations reveal a quasi-particle bandgap of 0.8eV of PbSe. STM results unveil that there are two types of defects on the surface, one is related the vacancies of Pb atoms and the other is the replacement of the absent Se atoms by Pb. Corresponding theoretical optimization confirms the structures of the defects. More importantly, both STS measurements and DFT calculations give evidence that the Pb vacancies move the Fermi energy inside the valence band and produce extra holes, leading to p-type characteristics of PbSe. Our work provides effective information for the future research of device performance based on PbSe films.

cond-mat.mtrl-sci↗

arXiv2023

Training-Free Energy Beamforming Assisted by Wireless Sensing

This paper studies the transmit energy beamforming in a multi-antenna wireless power transfer (WPT) system, in which an access point (AP) equipped with a uniform linear array (ULA) sends radio signals to wirelessly charge multiple single-antenna energy receivers (ERs). Different from conventional energy beamforming designs that require the AP to acquire the channel state information (CSI) via training and feedback, we propose a new training-free energy beamforming approach assisted by wireless radar sensing, which is implemented based on the following two-stage protocol. In the first stage, the AP performs wireless radar sensing to estimate the path gain and angle parameters of the ERs for constructing the corresponding CSI. In the second stage, the AP implements the transmit energy beamforming based on the constructed CSI to efficiently charge these ERs in a fair manner. Under this setup, first, we jointly optimize the sensing beamformers and duration in the first stage to minimize the sensing duration, while ensuring a given accuracy threshold for parameters estimation subject to the maximum transmit power constraint at the AP. Next, we optimize the energy beamformers in the second stage to maximize the minimum harvested energy by all ERs. In this approach, the estimation accuracy threshold for the first stage is properly designed to balance the resource allocation between the two stages for optimizing the ultimate energy harvesting performance. Finally, numerical results show that the proposed training-free energy beamforming design performs close to the performance upper bound with perfect CSI, and outperforms the benchmark schemes without such joint optimization and that with isotropic transmission.

cs.IT↗

arXiv2023

Generative Semantic Segmentation

We present Generative Semantic Segmentation (GSS), a generative learning approach for semantic segmentation. Uniquely, we cast semantic segmentation as an image-conditioned mask generation problem. This is achieved by replacing the conventional per-pixel discriminative learning with a latent prior learning process. Specifically, we model the variational posterior distribution of latent variables given the segmentation mask. To that end, the segmentation mask is expressed with a special type of image (dubbed as maskige). This posterior distribution allows to generate segmentation masks unconditionally. To achieve semantic segmentation on a given image, we further introduce a conditioning network. It is optimized by minimizing the divergence between the posterior distribution of maskige (i.e., segmentation masks) and the latent prior distribution of input training images. Extensive experiments on standard benchmarks show that our GSS can perform competitively to prior art alternatives in the standard semantic segmentation setting, whilst achieving a new state of the art in the more challenging cross-domain setting.

cs.CV↗

arXiv2023

The FlySpeech Audio-Visual Speaker Diarization System for MISP Challenge 2022

This paper describes the FlySpeech speaker diarization system submitted to the second \textbf{M}ultimodal \textbf{I}nformation Based \textbf{S}peech \textbf{P}rocessing~(\textbf{MISP}) Challenge held in ICASSP 2022. We develop an end-to-end audio-visual speaker diarization~(AVSD) system, which consists of a lip encoder, a speaker encoder, and an audio-visual decoder. Specifically, to mitigate the degradation of diarization performance caused by separate training, we jointly train the speaker encoder and the audio-visual decoder. In addition, we leverage the large-data pretrained speaker extractor to initialize the speaker encoder.

cs.SD↗

arXiv2023

Scale-aware Test-time Click Adaptation for Pulmonary Nodule and Mass Segmentation

Pulmonary nodules and masses are crucial imaging features in lung cancer screening that require careful management in clinical diagnosis. Despite the success of deep learning-based medical image segmentation, the robust performance on various sizes of lesions of nodule and mass is still challenging. In this paper, we propose a multi-scale neural network with scale-aware test-time adaptation to address this challenge. Specifically, we introduce an adaptive Scale-aware Test-time Click Adaptation method based on effortlessly obtainable lesion clicks as test-time cues to enhance segmentation performance, particularly for large lesions. The proposed method can be seamlessly integrated into existing networks. Extensive experiments on both open-source and in-house datasets consistently demonstrate the effectiveness of the proposed method over some CNN and Transformer-based segmentation methods. Our code is available at https://github.com/SplinterLi/SaTTCA

eess.IV↗

arXiv2023

Deep neural network improves the estimation of polygenic risk scores for breast cancer

Polygenic risk scores (PRS) estimate the genetic risk of an individual for a complex disease based on many genetic variants across the whole genome. In this study, we compared a series of computational models for estimation of breast cancer PRS. A deep neural network (DNN) was found to outperform alternative machine learning techniques and established statistical algorithms, including BLUP, BayesA and LDpred. In the test cohort with 50% prevalence, the Area Under the receiver operating characteristic Curve (AUC) were 67.4% for DNN, 64.2% for BLUP, 64.5% for BayesA, and 62.4% for LDpred. BLUP, BayesA, and LPpred all generated PRS that followed a normal distribution in the case population. However, the PRS generated by DNN in the case population followed a bi-modal distribution composed of two normal distributions with distinctly different means. This suggests that DNN was able to separate the case population into a high-genetic-risk case sub-population with an average PRS significantly higher than the control population and a normal-genetic-risk case sub-population with an average PRS similar to the control population. This allowed DNN to achieve 18.8% recall at 90% precision in the test cohort with 50% prevalence, which can be extrapolated to 65.4% recall at 20% precision in a general population with 12% prevalence. Interpretation of the DNN model identified salient variants that were assigned insignificant p-values by association studies, but were important for DNN prediction. These variants may be associated with the phenotype through non-linear relationships.

q-bio.QM↗

arXiv2023

Persistent Animal Identification Leveraging Non-Visual Markers

Our objective is to locate and provide a unique identifier for each mouse in a cluttered home-cage environment through time, as a precursor to automated behaviour recognition for biological research. This is a very challenging problem due to (i) the lack of distinguishing visual features for each mouse, and (ii) the close confines of the scene with constant occlusion, making standard visual tracking approaches unusable. However, a coarse estimate of each mouse's location is available from a unique RFID implant, so there is the potential to optimally combine information from (weak) tracking with coarse information on identity. To achieve our objective, we make the following key contributions: (a) the formulation of the object identification problem as an assignment problem (solved using Integer Linear Programming), and (b) a novel probabilistic model of the affinity between tracklets and RFID data. The latter is a crucial part of the model, as it provides a principled probabilistic treatment of object detections given coarse localisation. Our approach achieves 77% accuracy on this animal identification problem, and is able to reject spurious detections when the animals are hidden.

cs.CV↗