arXiv subjects

Zhe Liu

Publications and source records attributed to Zhe Liu.

At least 163 records · Page 9Linked to original sources

arXiv2025

Improving Adversarial Transferability on Vision Transformers via Forward Propagation Refinement

Vision Transformers (ViTs) have been widely applied in various computer vision and vision-language tasks. To gain insights into their robustness in practical scenarios, transferable adversarial examples on ViTs have been extensively studied. A typical approach to improving adversarial transferability is by refining the surrogate model. However, existing work on ViTs has restricted their surrogate refinement to backward propagation. In this work, we instead focus on Forward Propagation Refinement (FPR) and specifically refine two key modules of ViTs: attention maps and token embeddings. For attention maps, we propose Attention Map Diversification (AMD), which diversifies certain attention maps and also implicitly imposes beneficial gradient vanishing during backward propagation. For token embeddings, we propose Momentum Token Embedding (MTE), which accumulates historical token embeddings to stabilize the forward updates in both the Attention and MLP blocks. We conduct extensive experiments with adversarial examples transferred from ViTs to various CNNs and ViTs, demonstrating that our FPR outperforms the current best (backward) surrogate refinement by up to 7.0\% on average. We also validate its superiority against popular defenses and its compatibility with other transfer methods. Codes and appendix are available at https://github.com/RYC-98/FPR.

cs.CV↗

arXiv2025

DuSK: Faster Indirect Text Entry Supporting Out-Of-Vocabulary Words for Touchpads

Given the ubiquity of SmartTVs and head-mounted-display-based virtual environments, recent research has explored techniques to support eyes-free text entry using touchscreen devices. However, proposed techniques, leveraging lexicons, limit the user's ability to enter out-of-vocabulary words. In this paper, we investigate how to enter text while relying on unambiguous input to support out-of-vocabulary words. Through an iterative design approach, and after a careful investigation of actions that can be accurately and rapidly performed eyes-free, we devise DuSK, a {Du}al-handed, {S}troke-based, 1{K}eyboarding technique. In a controlled experiment, we show initial speeds of 10 WPM steadily increasing to 13~WPM with training. DuSK outperforms the common cursor-based text entry technique widely deployed in commercial SmartTVs (8 WPM) and is comparable to other eyes-free lexicon-based techniques, but with the added benefit of supporting out-of-vocabulary word input.

cs.HC↗

arXiv2025

Interview AI-ssistant: Designing for Real-Time Human-AI Collaboration in Interview Preparation and Execution

Recent advances in large language models (LLMs) offer unprecedented opportunities to enhance human-AI collaboration in qualitative research methods, including interviews. While interviews are highly valued for gathering deep, contextualized insights, interviewers often face significant cognitive challenges, such as real-time information processing, question adaptation, and rapport maintenance. My doctoral research introduces Interview AI-ssistant, a system designed for real-time interviewer-AI collaboration during both the preparation and execution phases. Through four interconnected studies, this research investigates the design of effective human-AI collaboration in interviewing contexts, beginning with a formative study of interviewers' needs, followed by a prototype development study focused on AI-assisted interview preparation, an experimental evaluation of real-time AI assistance during interviews, and a field study deploying the system in a real-world research setting. Beyond informing practical implementations of intelligent interview support systems, this work contributes to the Intelligent User Interfaces (IUI) community by advancing the understanding of human-AI collaborative interfaces in complex social tasks and establishing design guidelines for AI-enhanced qualitative research tools.

cs.HC↗

arXiv2025

Integration of a GaAs-based nanomechanical phase shifter with quantum-dot single-photon sources

We demonstrate a small-footprint electromechanical phase shifter with a compact active length of $10\:μ\text{m}$ fabricated on a suspended GaAs membrane, offering versatile integration with quantum-dot single-photon sources. The phase shifter is based on a slot-mode waveguide, whose slot width can be controlled by electrostatic forces, enabling a large effective refractive index change $Δn_\text{eff} > 0.1$. We observe up to 3$π$ phase modulation with 10.6 \text{V} applied bias, and a figure of merit $V_πL = 5.7\cdot 10^{-3} \text{V}\cdot \text{cm}$. Integration with a Mach-Zehnder interferometer (MZI) further allows routing of single photons with up to 24 \text{dB} extinction ratio at cryogenic temperatures. This device enables advanced manipulation of quantum emitters and the realization of reconfigurable quantum photonic integrated circuits.

quant-ph↗

arXiv2025

Controlling emitter-field coupling in waveguides with a nanomechanical phase shifter

The ability to control light-matter interfaces with solid-state photon emitters is a major requirement for the development of quantum photonic integrated circuits. We demonstrate controllable coupling between a quantum dot and an optical mode in a dielectric waveguide using a nano-opto-electro-mechanical phase shifter and a photonic crystal mirror. By controlling the phase, we induce a virtual displacement of the mirror that modifies the local density of states at the position of the emitter, thereby enhancing or suppressing spontaneous emission without an optical cavity. We observe a broadband tuning of the spontaneous emission rate and a modulation of the intensity emitted by the quantum dot in the waveguide. The method reported here could be employed to optimize the emitter-field interaction between quantum dots in-operando, by maximizing a single-photon source generation rate or adjusting its lifetime, as well as a characterization tool for the direct measurement of emitter-photon cooperativity.

quant-ph↗

arXiv2025

MAMBA4D: Efficient Long-Sequence Point Cloud Video Understanding with Disentangled Spatial-Temporal State Space Models

Point cloud videos can faithfully capture real-world spatial geometries and temporal dynamics, which are essential for enabling intelligent agents to understand the dynamically changing world. However, designing an effective 4D backbone remains challenging, mainly due to the irregular and unordered distribution of points and temporal inconsistencies across frames. Also, recent transformer-based 4D backbones commonly suffer from large computational costs due to their quadratic complexity, particularly for long video sequences. To address these challenges, we propose a novel point cloud video understanding backbone purely based on the State Space Models (SSMs). Specifically, we first disentangle space and time in 4D video sequences and then establish the spatio-temporal correlation with our designed Mamba blocks. The Intra-frame Spatial Mamba module is developed to encode locally similar geometric structures within a certain temporal stride. Subsequently, locally correlated tokens are delivered to the Inter-frame Temporal Mamba module, which integrates long-term point features across the entire video with linear complexity. Our proposed Mamba4d achieves competitive performance on the MSR-Action3D action recognition (+10.4% accuracy), HOI4D action segmentation (+0.7 F1 Score), and Synthia4D semantic segmentation (+0.19 mIoU) datasets. Especially, for long video sequences, our method has a significant efficiency improvement with 87.5% GPU memory reduction and 5.36 times speed-up. Codes will be released at https://github.com/IRMVLab/Mamba4D.

cs.CV↗

arXiv2025

Exploring the Role of Explicit Temporal Modeling in Multimodal Large Language Models for Video Understanding

Applying Multimodal Large Language Models (MLLMs) to video understanding presents significant challenges due to the need to model temporal relations across frames. Existing approaches adopt either implicit temporal modeling, relying solely on the LLM decoder, or explicit temporal modeling, employing auxiliary temporal encoders. To investigate this debate between the two paradigms, we propose the Stackable Temporal Encoder (STE). STE enables flexible explicit temporal modeling with adjustable temporal receptive fields and token compression ratios. Using STE, we systematically compare implicit and explicit temporal modeling across dimensions such as overall performance, token compression effectiveness, and temporal-specific understanding. We also explore STE's design considerations and broader impacts as a plug-in module and in image modalities. Our findings emphasize the critical role of explicit temporal modeling, providing actionable insights to advance video MLLMs.

cs.CV↗

arXiv2025

FLAME: Learning to Navigate with Multimodal LLM in Urban Environments

Large Language Models (LLMs) have demonstrated potential in Vision-and-Language Navigation (VLN) tasks, yet current applications face challenges. While LLMs excel in general conversation scenarios, they struggle with specialized navigation tasks, yielding suboptimal performance compared to specialized VLN models. We introduce FLAME (FLAMingo-Architected Embodied Agent), a novel Multimodal LLM-based agent and architecture designed for urban VLN tasks that efficiently handles multiple observations. Our approach implements a three-phase tuning technique for effective adaptation to navigation tasks, including single perception tuning for street view description, multiple perception tuning for route summarization, and end-to-end training on VLN datasets. The augmented datasets are synthesized automatically. Experimental results demonstrate FLAME's superiority over existing methods, surpassing state-of-the-art methods by a 7.3% increase in task completion on Touchdown dataset. This work showcases the potential of Multimodal LLMs (MLLMs) in complex navigation tasks, representing an advancement towards applications of MLLMs in the field of embodied intelligence.

cs.CV↗

arXiv2024

Planning from Imagination: Episodic Simulation and Episodic Memory for Vision-and-Language Navigation

Humans navigate unfamiliar environments using episodic simulation and episodic memory, which facilitate a deeper understanding of the complex relationships between environments and objects. Developing an imaginative memory system inspired by human mechanisms can enhance the navigation performance of embodied agents in unseen environments. However, existing Vision-and-Language Navigation (VLN) agents lack a memory mechanism of this kind. To address this, we propose a novel architecture that equips agents with a reality-imagination hybrid memory system. This system enables agents to maintain and expand their memory through both imaginative mechanisms and navigation actions. Additionally, we design tailored pre-training tasks to develop the agent's imaginative capabilities. Our agent can imagine high-fidelity RGB images for future scenes, achieving state-of-the-art result in Success rate weighted by Path Length (SPL).

cs.CV↗

arXiv2024

Improving Integrated Gradient-based Transferable Adversarial Examples by Refining the Integration Path

Transferable adversarial examples are known to cause threats in practical, black-box attack scenarios. A notable approach to improving transferability is using integrated gradients (IG), originally developed for model interpretability. In this paper, we find that existing IG-based attacks have limited transferability due to their naive adoption of IG in model interpretability. To address this limitation, we focus on the IG integration path and refine it in three aspects: multiplicity, monotonicity, and diversity, supported by theoretical analyses. We propose the Multiple Monotonic Diversified Integrated Gradients (MuMoDIG) attack, which can generate highly transferable adversarial examples on different CNN and ViT models and defenses. Experiments validate that MuMoDIG outperforms the latest IG-based attack by up to 37.3\% and other state-of-the-art attacks by 8.4\%. In general, our study reveals that migrating established techniques to improve transferability may require non-trivial efforts. Code is available at \url{https://github.com/RYC-98/MuMoDIG}.

cs.CR↗

arXiv2024

DR-Encoder: Encode Low-rank Gradients with Random Prior for Large Language Models Differentially Privately

The emergence of the Large Language Model (LLM) has shown their superiority in a wide range of disciplines, including language understanding and translation, relational logic reasoning, and even partial differential equations solving. The transformer is the pervasive backbone architecture for the foundation model construction. It is vital to research how to adjust the Transformer architecture to achieve an end-to-end privacy guarantee in LLM fine-tuning. In this paper, we investigate three potential information leakage during a federated fine-tuning procedure for LLM (FedLLM). Based on the potential information leakage, we provide an end-to-end privacy guarantee solution for FedLLM by inserting two-stage randomness. The first stage is to train a gradient auto-encoder with a Gaussian random prior based on the statistical information of the gradients generated by local clients. The second stage is to fine-tune the overall LLM with a differential privacy guarantee by adopting appropriate Gaussian noises. We show the efficiency and accuracy gains of our proposed method with several foundation models and two popular evaluation benchmarks. Furthermore, we present a comprehensive privacy analysis with Gaussian Differential Privacy (GDP) and Renyi Differential Privacy (RDP).

cs.LG↗

arXiv2024

A Roadmap for Software Testing in Open Collaborative Development Environments

Amidst the ever-expanding digital sphere, the evolution of the Internet has not only fostered an atmosphere of information transparency and sharing but has also sparked a revolution in software development practices. The distributed nature of open collaborative development, along with its diverse contributors and rapid iterations, presents new challenges for ensuring software quality. This paper offers a comprehensive review and analysis of recent advancements in software quality assurance within open collaborative development environments. Our examination covers various aspects, including process management, personnel dynamics, and technological advancements, providing valuable insights into effective approaches for maintaining software quality in such collaborative settings. Furthermore, we delve into the challenges and opportunities arising from emerging technologies such as LLMs and the AI model-centric development paradigm. By addressing these topics, our study contributes to a deeper understanding of software quality assurance in open collaborative environments and lays the groundwork for future exploration and innovation.

cs.SE↗

arXiv2024

Machine Learning Co-pilot for Screening of Organic Molecular Additives for Perovskite Solar Cells

Machine learning (ML) has been extensively employed in planar perovskite photovoltaics to screen effective organic molecular additives, while encountering predictive biases for novel materials due to small datasets and reliance on predefined descriptors. Present work thus proposes an effective approach, Co-Pilot for Perovskite Additive Screener (Co-PAS), an ML-driven framework designed to accelerate additive screening for perovskite solar cells (PSCs). Co-PAS overcomes predictive biases by integrating the Molecular Scaffold Classifier (MSC) for scaffold-based pre-screening and utilizing Junction Tree Variational Autoencoder (JTVAE) latent vectors to enhance molecular structure representation, thereby enhancing the accuracy of power conversion efficiency (PCE) predictions. Leveraging Co-PAS, we integrate domain knowledge to screen an extensive dataset of 250,000 molecules from PubChem, prioritizing candidates based on predicted PCE values and key molecular properties such as donor number, dipole moment, and hydrogen bond acceptor count. This workflow leads to the identification of several promising passivating molecules, including the novel Boc-L-threonine N-hydroxysuccinimide ester (BTN), which, to our knowledge, has not been explored as an additive in PSCs and achieves a device PCE of 25.20%. Our results underscore the potential of Co-PAS in advancing additive discovery for high-performance PSCs.

cs.LG↗

arXiv2024

Trusted Unified Feature-Neighborhood Dynamics for Multi-View Classification

Multi-view classification (MVC) faces inherent challenges due to domain gaps and inconsistencies across different views, often resulting in uncertainties during the fusion process. While Evidential Deep Learning (EDL) has been effective in addressing view uncertainty, existing methods predominantly rely on the Dempster-Shafer combination rule, which is sensitive to conflicting evidence and often neglects the critical role of neighborhood structures within multi-view data. To address these limitations, we propose a Trusted Unified Feature-NEighborhood Dynamics (TUNED) model for robust MVC. This method effectively integrates local and global feature-neighborhood (F-N) structures for robust decision-making. Specifically, we begin by extracting local F-N structures within each view. To further mitigate potential uncertainties and conflicts in multi-view fusion, we employ a selective Markov random field that adaptively manages cross-view neighborhood dependencies. Additionally, we employ a shared parameterized evidence extractor that learns global consensus conditioned on local F-N structures, thereby enhancing the global integration of multi-view features. Experiments on benchmark datasets show that our method improves accuracy and robustness over existing approaches, particularly in scenarios with high uncertainty and conflicting views. The code will be made available at https://github.com/JethroJames/TUNED.

cs.CV↗

arXiv2024

Compositional Zero-Shot Learning with Contextualized Cues and Adaptive Contrastive Training

Compositional Zero-Shot Learning (CZSL) aims to recognize unseen combinations of seen attributes and objects. Current CLIP-based methods in CZSL, despite their advancements, often fail to effectively understand and link the attributes and objects due to inherent limitations in CLIP's pretraining mechanisms. To address these shortcomings, this paper introduces a novel framework, Understanding and Linking Attributes and Objects (ULAO) in CZSL, which comprises two innovative modules. The Understanding Attributes and Objects (UAO) module improves primitive understanding by sequential primitive prediction and leveraging recognized objects as contextual hints for attribute classification. Concurrently, the Linking Attributes and Objects (LAO) module improves the attribute-object linkage understanding through a new contrastive learning strategy that incorporates tailored hard negative generation and adaptive loss adjustments. We demonstrate our model's superiority by showcasing its state-of-the-art performance across three benchmark datasets in both Closed-World (CW) and Open-World (OW) scenarios.

cs.CV↗

arXiv2024

Effective Text Adaptation for LLM-based ASR through Soft Prompt Fine-Tuning

The advent of Large Language Models (LLM) has reformed the Automatic Speech Recognition (ASR). Prompting LLM with audio embeddings to generate transcriptions becomes the new state-of-the-art ASR. Despite LLMs being trained with an extensive amount of text corpora, high-quality domain-specific text data can still significantly enhance ASR performance on domain adaptation tasks. Although LLM-based ASR can naturally incorporate more text corpora by fine-tuning the LLM decoder, fine-tuning such ASR on text-only data without paired prompts may diminish the effectiveness of domain-specific knowledge. To mitigate this issue, we propose a two-step soft prompt fine-tuning strategy that enhances domain-specific text adaptation. Experimental results show that text adaptation with our proposed method achieved a relative up to 9% Word Error Rate (WER) reduction and up to 18% Entity Error Rate (EER) reduction on the target domain compared to the baseline ASR. Combining this with domain-specific Language Model (LM) fusion can further improve the EER by a relative 2-5%

cs.CL↗

arXiv2024

Seeing is Believing: Vision-driven Non-crash Functional Bug Detection for Mobile Apps

Mobile app GUI (Graphical User Interface) pages now contain rich visual information, with the visual semantics of each page helping users understand the application logic. However, these complex visual and functional logic present new challenges to software testing. Existing automated GUI testing methods, constrained by the lack of reliable testing oracles, are limited to detecting crash bugs with obvious abnormal signals. Consequently, many non-crash functional bugs, ranging from unexpected behaviors to logical errors, often evade detection by current techniques. While these non-crash functional bugs can exhibit visual cues that serve as potential testing oracles, they often entail a sequence of screenshots, and detecting them necessitates an understanding of the operational logic among GUI page transitions, which is challenging traditional techniques. Considering the remarkable performance of Multimodal Large Language Models (MLLM) in visual and language understanding, this paper proposes Trident, a novel vision-driven, multi-agent collaborative automated GUI testing approach for detecting non-crash functional bugs. It comprises three agents: Explorer, Monitor, and Detector, to guide the exploration, oversee the testing progress, and spot issues. We also address several challenges, i.e., align visual and textual information for MLLM input, achieve functionality-oriented exploration, and infer test oracles for non-crash bugs, to enhance the performance of functionality bug detection. We evaluate Trident on 590 non-crash bugs and compare it with 12 baselines, it can achieve more than 14%-112% and 108%-147% boost in average recall and precision compared with the best baseline. The ablation study further proves the contribution of each module. Moreover, Trident identifies 43 new bugs on Google Play, of which 31 have been fixed.

cs.SE↗

arXiv2024

PG-SLAM: Photo-realistic and Geometry-aware RGB-D SLAM in Dynamic Environments

Simultaneous localization and mapping (SLAM) has achieved impressive performance in static environments. However, SLAM in dynamic environments remains an open question. Many methods directly filter out dynamic objects, resulting in incomplete scene reconstruction and limited accuracy of camera localization. The other works express dynamic objects by point clouds, sparse joints, or coarse meshes, which fails to provide a photo-realistic representation. To overcome the above limitations, we propose a photo-realistic and geometry-aware RGB-D SLAM method by extending Gaussian splatting. Our method is composed of three main modules to 1) map the dynamic foreground including non-rigid humans and rigid items, 2) reconstruct the static background, and 3) localize the camera. To map the foreground, we focus on modeling the deformations and/or motions. We consider the shape priors of humans and exploit geometric and appearance constraints of humans and items. For background mapping, we design an optimization strategy between neighboring local maps by integrating appearance constraint into geometric alignment. As to camera localization, we leverage both static background and dynamic foreground to increase the observations for noise compensation. We explore the geometric and appearance constraints by associating 3D Gaussians with 2D optical flows and pixel patches. Experiments on various real-world datasets demonstrate that our method outperforms state-of-the-art approaches in terms of camera localization and scene representation. Source codes will be publicly available upon paper acceptance.

cs.RO↗