arXiv subjects

Cheng Zhang

Publications and source records attributed to Cheng Zhang.

At least 253 records · Page 14Linked to original sources

arXiv2024

Generalized Maximum Likelihood Estimation for Perspective-n-Point Problem

The Perspective-n-Point (PnP) problem has been widely studied in the literature and applied in various vision-based pose estimation scenarios. However, existing methods ignore the anisotropy uncertainty of observations, as demonstrated in several real-world datasets in this paper. This oversight may lead to suboptimal and inaccurate estimation, particularly in the presence of noisy observations. To this end, we propose a generalized maximum likelihood PnP solver, named GMLPnP, that minimizes the determinant criterion by iterating the GLS procedure to estimate the pose and uncertainty simultaneously. Further, the proposed method is decoupled from the camera model. Results of synthetic and real experiments show that our method achieves better accuracy in common pose estimation scenarios, GMLPnP improves rotation/translation accuracy by 4.7%/2.0% on TUM-RGBD and 18.6%/18.4% on KITTI-360 dataset compared to the best baseline. It is more accurate under very noisy observations in a vision-based UAV localization task, outperforming the best baseline by 34.4% in translation estimation accuracy.

cs.CV↗

arXiv2024

Crystal-symmetry-paired spin-valley locking in a layered room-temperature antiferromagnet

Recent theoretical efforts predicted a type of unconventional antiferromagnet characterized by the crystal symmetry C (rotation or mirror), which connects antiferromagnetic sublattices in real space and simultaneously couples spin and momentum in reciprocal space. This results in a unique C-paired spin-valley locking (SVL) and corresponding novel properties such as piezomagnetism and noncollinear spin current even without spin-orbit coupling. However, the unconventional antiferromagnets reported thus far are not layered materials, limiting their potential in spintronic applications. Additionally, they do not meet the necessary symmetry requirements for nonrelativistic spin current. Here, we report the realization of C-paired SVL in a layered room-temperature antiferromagnetic compound, Rb1-δV2Te2O. Spin resolved photoemission measurements directly demonstrate the opposite spin splitting between C-paired valleys. Quasi-particle interference patterns reveal the suppression of inter-valley scattering due to the spin selection rules, as a direct consequence of C-paired SVL. All these experiments are well consistent with the results obtained from first-principles calculations. Our observations represent the first realization of layered antiferromagnets with C-paired SVL, enabling both the advantages of layered materials and possible control through crystal symmetry manipulation. These results hold significant promise and broad implications for advancements in magnetism, electronics, and information technology.

cond-mat.str-el↗

arXiv2024

MunchSonic: Tracking Fine-grained Dietary Actions through Active Acoustic Sensing on Eyeglasses

We introduce MunchSonic, an AI-powered active acoustic sensing system integrated into eyeglasses to track fine-grained dietary actions. MunchSonic emits inaudible ultrasonic waves from the eyeglass frame, with the reflected signals capturing detailed positions and movements of body parts, including the mouth, jaw, arms, and hands involved in eating. These signals are processed by a deep learning pipeline to classify six actions: hand-to-mouth movements for food intake, chewing, drinking, talking, face-hand touching, and other activities (null). In an unconstrained study with 12 participants, MunchSonic achieved a 93.5% macro F1-score in a user-independent evaluation with a 2-second resolution in tracking these actions, also demonstrating its effectiveness in tracking eating episodes and food intake frequency within those episodes.

cs.HC↗

arXiv2024

EchoGuide: Active Acoustic Guidance for LLM-Based Eating Event Analysis from Egocentric Videos

Self-recording eating behaviors is a step towards a healthy lifestyle recommended by many health professionals. However, the current practice of manually recording eating activities using paper records or smartphone apps is often unsustainable and inaccurate. Smart glasses have emerged as a promising wearable form factor for tracking eating behaviors, but existing systems primarily identify when eating occurs without capturing details of the eating activities (E.g., what is being eaten). In this paper, we present EchoGuide, an application and system pipeline that leverages low-power active acoustic sensing to guide head-mounted cameras to capture egocentric videos, enabling efficient and detailed analysis of eating activities. By combining active acoustic sensing for eating detection with video captioning models and large-scale language models for retrieval augmentation, EchoGuide intelligently clips and analyzes videos to create concise, relevant activity records on eating. We evaluated EchoGuide with 9 participants in naturalistic settings involving eating activities, demonstrating high-quality summarization and significant reductions in video data needed, paving the way for practical, scalable eating activity tracking.

cs.HC↗

arXiv2024

An Empirical Study and Analysis of Text-to-Image Generation Using Large Language Model-Powered Textual Representation

One critical prerequisite for faithful text-to-image generation is the accurate understanding of text inputs. Existing methods leverage the text encoder of the CLIP model to represent input prompts. However, the pre-trained CLIP model can merely encode English with a maximum token length of 77. Moreover, the model capacity of the text encoder from CLIP is relatively limited compared to Large Language Models (LLMs), which offer multilingual input, accommodate longer context, and achieve superior text representation. In this paper, we investigate LLMs as the text encoder to improve the language understanding in text-to-image generation. Unfortunately, training text-to-image generative model with LLMs from scratch demands significant computational resources and data. To this end, we introduce a three-stage training pipeline that effectively and efficiently integrates the existing text-to-image model with LLMs. Specifically, we propose a lightweight adapter that enables fast training of the text-to-image model using the textual representations from LLMs. Extensive experiments demonstrate that our model supports not only multilingual but also longer input context with superior image generation quality.

cs.CV↗

arXiv2024

Generalizable Human Gaussians for Sparse View Synthesis

Recent progress in neural rendering has brought forth pioneering methods, such as NeRF and Gaussian Splatting, which revolutionize view rendering across various domains like AR/VR, gaming, and content creation. While these methods excel at interpolating {\em within the training data}, the challenge of generalizing to new scenes and objects from very sparse views persists. Specifically, modeling 3D humans from sparse views presents formidable hurdles due to the inherent complexity of human geometry, resulting in inaccurate reconstructions of geometry and textures. To tackle this challenge, this paper leverages recent advancements in Gaussian Splatting and introduces a new method to learn generalizable human Gaussians that allows photorealistic and accurate view-rendering of a new human subject from a limited set of sparse views in a feed-forward manner. A pivotal innovation of our approach involves reformulating the learning of 3D Gaussian parameters into a regression process defined on the 2D UV space of a human template, which allows leveraging the strong geometry prior and the advantages of 2D convolutions. In addition, a multi-scaffold is proposed to effectively represent the offset details. Our method outperforms recent methods on both within-dataset generalization as well as cross-dataset generalization settings.

cs.CV↗

arXiv2024

Unlocking the Global Synergies in Low-Rank Adapters

Low-rank Adaption (LoRA) has been the de-facto parameter-efficient fine-tuning technique for large language models. We present HeteroLoRA, a light-weight search algorithm that leverages zero-cost proxies to allocate the limited LoRA trainable parameters across the model for better fine-tuned performance. In addition to the allocation for the standard LoRA-adapted models, we also demonstrate the efficacy of HeteroLoRA by performing the allocation in a more challenging search space that includes LoRA modules and LoRA-adapted shortcut connections. Experiments show that HeteroLoRA enables improvements in model performance given the same parameter budge. For example, on MRPC, we see an improvement of 1.6% in accuracy with similar training parameter budget. We will open-source our algorithm once the paper is accepted.

cs.LG↗

arXiv2024

Optimised Grouped-Query Attention Mechanism for Transformers

Grouped-query attention (GQA) has been widely adopted in LLMs to mitigate the complexity of multi-head attention (MHA). To transform an MHA to a GQA, neighbour queries in MHA are evenly split into groups where each group shares the value and key layers. In this work, we propose AsymGQA, an activation-informed approach to asymmetrically grouping an MHA to a GQA for better model performance. Our AsymGQA outperforms the GQA within the same model size budget. For example, AsymGQA LLaMA-2-7B has an accuracy increase of 7.5% on MMLU compared to neighbour grouping. Our approach addresses the GQA's trade-off problem between model performance and hardware efficiency.

cs.LG↗

arXiv2024

You Only Need One Color Space: An Efficient Network for Low-light Image Enhancement

Low-Light Image Enhancement (LLIE) task tends to restore the details and visual information from corrupted low-light images. Most existing methods learn the mapping function between low/normal-light images by Deep Neural Networks (DNNs) on sRGB and HSV color space. Nevertheless, enhancement involves amplifying image signals, and applying these color spaces to low-light images with a low signal-to-noise ratio can introduce sensitivity and instability into the enhancement process. Consequently, this results in the presence of color artifacts and brightness artifacts in the enhanced images. To alleviate this problem, we propose a novel trainable color space, named Horizontal/Vertical-Intensity (HVI). It not only decouples brightness and color from RGB channels to mitigate the instability during enhancement but also adapts to low-light images in different illumination ranges due to the trainable parameters. Further, we design a novel Color and Intensity Decoupling Network (CIDNet) with two branches dedicated to processing the decoupled image brightness and color in the HVI space. Within CIDNet, we introduce the Lightweight Cross-Attention (LCA) module to facilitate interaction between image structure and content information in both branches, while also suppressing noise in low-light images. Finally, we conducted 22 quantitative and qualitative experiments to show that the proposed CIDNet outperforms the state-of-the-art methods on 11 datasets. The code is available at https://github.com/Fediory/HVI-CIDNet.

cs.CV↗

arXiv2024

Photon induced proton and anti-proton pair production with ultraperipheral heavy ion collisions at RHIC

We investigate proton-antiproton ($p\bar{p}$) pair production via photon-photon fusion in the ultra-peripheral collisions at RHIC, employing a joint impact parameter and transverse momentum dependent formalism. We consider proton exchange, $s$-channel resonance and hand-bag mechanisms, predicting differential distributions of $p\bar p$ production. Our theoretical predictions can be tested against future measurements at RHIC, to enhance our understanding of photon-photon interactions in strong electromagnetic fields.

hep-ph↗

arXiv2024

HASS: Hardware-Aware Sparsity Search for Dataflow DNN Accelerator

Deep Neural Networks (DNNs) excel in learning hierarchical representations from raw data, such as images, audio, and text. To compute these DNN models with high performance and energy efficiency, these models are usually deployed onto customized hardware accelerators. Among various accelerator designs, dataflow architecture has shown promising performance due to its layer-pipelined structure and its scalability in data parallelism. Exploiting weights and activations sparsity can further enhance memory storage and computation efficiency. However, existing approaches focus on exploiting sparsity in non-dataflow accelerators, which cannot be applied onto dataflow accelerators because of the large hardware design space introduced. As such, this could miss opportunities to find an optimal combination of sparsity features and hardware designs. In this paper, we propose a novel approach to exploit unstructured weights and activations sparsity for dataflow accelerators, using software and hardware co-optimization. We propose a Hardware-Aware Sparsity Search (HASS) to systematically determine an efficient sparsity solution for dataflow accelerators. Over a set of models, we achieve an efficiency improvement ranging from 1.3$\times$ to 4.2$\times$ compared to existing sparse designs, which are either non-dataflow or non-hardware-aware. Particularly, the throughput of MobileNetV3 can be optimized to 4895 images per second. HASS is open-source: \url{https://github.com/Yu-Zhewen/HASS}

cs.AR↗

arXiv2024

Training of Physical Neural Networks

Physical neural networks (PNNs) are a class of neural-like networks that leverage the properties of physical systems to perform computation. While PNNs are so far a niche research area with small-scale laboratory demonstrations, they are arguably one of the most underappreciated important opportunities in modern AI. Could we train AI models 1000x larger than current ones? Could we do this and also have them perform inference locally and privately on edge devices, such as smartphones or sensors? Research over the past few years has shown that the answer to all these questions is likely "yes, with enough research": PNNs could one day radically change what is possible and practical for AI systems. To do this will however require rethinking both how AI models work, and how they are trained - primarily by considering the problems through the constraints of the underlying hardware physics. To train PNNs at large scale, many methods including backpropagation-based and backpropagation-free approaches are now being explored. These methods have various trade-offs, and so far no method has been shown to scale to the same scale and performance as the backpropagation algorithm widely used in deep learning today. However, this is rapidly changing, and a diverse ecosystem of training techniques provides clues for how PNNs may one day be utilized to create both more efficient realizations of current-scale AI models, and to enable unprecedented-scale models.

physics.app-ph↗

arXiv2024

Towards Causal Foundation Model: on Duality between Causal Inference and Attention

Foundation models have brought changes to the landscape of machine learning, demonstrating sparks of human-level intelligence across a diverse array of tasks. However, a gap persists in complex tasks such as causal inference, primarily due to challenges associated with intricate reasoning steps and high numerical precision requirements. In this work, we take a first step towards building causally-aware foundation models for treatment effect estimations. We propose a novel, theoretically justified method called Causal Inference with Attention (CInA), which utilizes multiple unlabeled datasets to perform self-supervised causal learning, and subsequently enables zero-shot causal inference on unseen tasks with new data. This is based on our theoretical results that demonstrate the primal-dual connection between optimal covariate balancing and self-attention, facilitating zero-shot causal inference through the final layer of a trained transformer-type architecture. We demonstrate empirically that CInA effectively generalizes to out-of-distribution datasets and various real-world datasets, matching or even surpassing traditional per-dataset methodologies. These results provide compelling evidence that our method has the potential to serve as a stepping stone for the development of causal foundation models.

cs.LG↗

arXiv2024

Learning Adaptive Fusion Bank for Multi-modal Salient Object Detection

Multi-modal salient object detection (MSOD) aims to boost saliency detection performance by integrating visible sources with depth or thermal infrared ones. Existing methods generally design different fusion schemes to handle certain issues or challenges. Although these fusion schemes are effective at addressing specific issues or challenges, they may struggle to handle multiple complex challenges simultaneously. To solve this problem, we propose a novel adaptive fusion bank that makes full use of the complementary benefits from a set of basic fusion schemes to handle different challenges simultaneously for robust MSOD. We focus on handling five major challenges in MSOD, namely center bias, scale variation, image clutter, low illumination, and thermal crossover or depth ambiguity. The fusion bank proposed consists of five representative fusion schemes, which are specifically designed based on the characteristics of each challenge, respectively. The bank is scalable, and more fusion schemes could be incorporated into the bank for more challenges. To adaptively select the appropriate fusion scheme for multi-modal input, we introduce an adaptive ensemble module that forms the adaptive fusion bank, which is embedded into hierarchical layers for sufficient fusion of different source data. Moreover, we design an indirect interactive guidance module to accurately detect salient hollow objects via the skip integration of high-level semantic information and low-level spatial details. Extensive experiments on three RGBT datasets and seven RGBD datasets demonstrate that the proposed method achieves the outstanding performance compared to the state-of-the-art methods. The code and results are available at https://github.com/Angknpng/LAFB.

cs.CV↗

arXiv2024

LQER: Low-Rank Quantization Error Reconstruction for LLMs

Post-training quantization of Large Language Models (LLMs) is challenging. In this work, we introduce Low-rank Quantization Error Reduction (LQER), which combines quantization and low-rank approximation to recover the model capability. LQER leverages an activation-induced scale matrix to drive the singular value distribution of quantization error towards a desirable distribution, which enables nearly-lossless W4A8 quantization on various LLMs and downstream tasks without the need for knowledge distillation, grid search, or gradient-base iterative optimization. Unlike existing methods, the computation pattern of LQER eliminates the need for specialized Scatter and Gather processes to collect high-precision weights from irregular memory locations. Our W4A8 LLMs achieve near-lossless performance on six popular downstream tasks, while using 1.36$\times$ fewer hardware resources than the leading state-of-the-art method. We open-source our framework at https://github.com/ChengZhang-98/lqer

cs.LG↗

arXiv2024

Kernel Semi-Implicit Variational Inference

Semi-implicit variational inference (SIVI) extends traditional variational families with semi-implicit distributions defined in a hierarchical manner. Due to the intractable densities of semi-implicit distributions, classical SIVI often resorts to surrogates of evidence lower bound (ELBO) that would introduce biases for training. A recent advancement in SIVI, named SIVI-SM, utilizes an alternative score matching objective made tractable via a minimax formulation, albeit requiring an additional lower-level optimization. In this paper, we propose kernel SIVI (KSIVI), a variant of SIVI-SM that eliminates the need for lower-level optimization through kernel tricks. Specifically, we show that when optimizing over a reproducing kernel Hilbert space (RKHS), the lower-level problem has an explicit solution. This way, the upper-level objective becomes the kernel Stein discrepancy (KSD), which is readily computable for stochastic gradient descent due to the hierarchical structure of semi-implicit variational distributions. An upper bound for the variance of the Monte Carlo gradient estimators of the KSD objective is derived, which allows us to establish novel convergence guarantees of KSIVI. We demonstrate the effectiveness and efficiency of KSIVI on both synthetic distributions and a variety of real data Bayesian inference tasks.

stat.ML↗

arXiv2024

Memory-assisted measurement-device-independent quantum secret sharing

Measurement-device-independent quantum secret sharing (MDI-QSS) can eliminate all the security loopholes associated with imperfect measurement devices and greatly enhance QS's security under practical experimental condition. MDI-QSS requires each communication user to send single photon to the measurement party for the coincident measurement. However, the unsynchronization of the transmitted photons greatly limits MDI-QSS's practical performance.In the paper, we propose a high-efficient quantum memory (QM)-assisted MDI-QSS protocol, which employs the QM-assisted synchronization of three heralded single-photon sources to efficiently generate three simultaneous single-photon states. The QM constructed with all-optical, polarization-insensitive storage loop has superior performance in terms of bandwidth, storage efficiency, and noise resistance, and is feasible under current experiment conditions. Combining with the decoy-state method, we perform the numerical simulation of the secure key rate in the symmetric model without considering the finite-size effect. The simulation results show that our QM-assisted MDI-QSS protocol exhibit largely improved secure key rate and maximal photon transmission distance compared with all existing MDI-QSS protocols without QM. Our protocol provides a promising way for implementing the high-efficient long-distance MDI-QSS in the near future.

quant-ph↗

arXiv2024

Symmetry quantification and segmentation in STEM imaging through Zernike moments

We present a method using Zernike moments for quantifying rotational and reflectional symmetries in scanning transmission electron microscopy (STEM) images, aimed at improving structural analysis of materials at the atomic scale. This technique is effective against common imaging noises and is potentially suited for low-dose imaging and identifying quantum defects. We showcase its utility in the unsupervised segmentation of polytypes in a twisted bilayer TaS$_2$, enabling accurate differentiation of structural phases and monitoring transitions caused by electron beam effects. This approach enhances the analysis of structural variations in crystalline materials, marking a notable advancement in the characterization of structures in materials science.

cond-mat.mtrl-sci↗