Search

For elliptic curves over rationals, there are a well-known conjecture of Sato-Tate and a new computational guided murmuration phenomenon, for which the abelian Hasse-Weil zeta functions are used. In this paper, we show that both the murmurations and the Sato-Tate conjecture stand equally well for non-abelian high rank zeta functions of the p-reductions of elliptic curves over rationals., Comment: 18 pages, 7 figures

Road segmentation is a critical task for autonomous driving systems, requiring accurate and robust methods to classify road surfaces from various environmental data. Our work introduces an innovative approach that integrates LiDAR point cloud data, visual image, and relative depth maps derived from images. The integration of multiple data sources in road segmentation presents both opportunities and challenges. One of the primary challenges is the scarcity of large-scale, accurately labeled datasets that are necessary for training robust deep learning models. To address this, we have developed the [UdeerLID+] framework under a semi-supervised learning paradigm. Experiments results on KITTI datasets validate the superior performance.

Let $(V(u),\, u\in \mathcal{T})$ be a (supercritical) branching random walk and $(\eta_u,\,u\in \mathcal{T})$ be marks on the vertices of the tree, distributed in an i.i.d.\ fashion. Following Aldous and Bandyopadhyay \cite{AB05}, for each infinite ray $\xi$ of the tree, we associate the {\it discounted tree sum} $D(\xi)$ which is the sum of the $e^{-V(u)}\eta_u$ taken along the ray. The paper deals with the finiteness of $\sup_\xi D(\xi)$. To this end, we study the extreme behaviour of the local time processes of the paths $(V(u),\,u\in \xi)$. It answers a question of Nicolas Curien, and partially solves Open Problem 31 of Aldous and Bandyopadhyay \cite{AB05}. We also present several open questions.

We introduce a novel MV-DETR pipeline which is effective while efficient transformer based detection method. Given input RGBD data, we notice that there are super strong pretraining weights for RGB data while less effective works for depth related data. First and foremost , we argue that geometry and texture cues are both of vital importance while could be encoded separately. Secondly, we find that visual texture feature is relatively hard to extract compared with geometry feature in 3d space. Unfortunately, single RGBD dataset with thousands of data is not enough for training an discriminating filter for visual texture feature extraction. Last but certainly not the least, we designed a lightweight VG module consists of a visual textual encoder, a geometry encoder and a VG connector. Compared with previous state of the art works like V-DETR, gains from pretrained visual encoder could be seen. Extensive experiments on ScanNetV2 dataset shows the effectiveness of our method. It is worth mentioned that our method achieve 78\% AP which create new state of the art on ScanNetv2 benchmark.

In the context of smart city development, mobile sensing emerges as a cost-effective alternative to fixed sensing for on-street parking detection. However, its practicality is often challenged by the inherent accuracy limitations arising from detection intervals. This paper introduces a novel Dynamic Gap Reduction Algorithm (DGRA), which is a crowdsensing-based approach aimed at addressing this question through parking detection data collected by sensors on moving vehicles. The algorithm's efficacy is validated through real drive tests and simulations. We also present a Driver-Side and Traffic-Based Model (DSTBM), which incorporates drivers' parking decisions and traffic conditions to evaluate DGRA's performance. Results highlight DGRA's significant potential in reducing the mobile sensing accuracy gap, marking a step forward in efficient urban parking management.

The human brain has the capability to associate the unknown person's voice and face by leveraging their general relationship, referred to as ``cross-modal speaker verification''. This task poses significant challenges due to the complex relationship between the modalities. In this paper, we propose a ``Multi-stage Face-voice Association Learning with Keynote Speaker Diarization''~(MFV-KSD) framework. MFV-KSD contains a keynote speaker diarization front-end to effectively address the noisy speech inputs issue. To balance and enhance the intra-modal feature learning and inter-modal correlation understanding, MFV-KSD utilizes a novel three-stage training strategy. Our experimental results demonstrated robust performance, achieving the first rank in the 2024 Face-voice Association in Multilingual Environments (FAME) challenge with an overall Equal Error Rate (EER) of 19.9%. Details can be found in https://github.com/TaoRuijie/MFV-KSD.

We study the random walk on dynamical percolation of $\mathbb{Z}^d$ (resp., the two-dimensional triangular lattice $\mathcal{T}$), where each edge (resp., each site) can be either open or closed, refreshing its status at rate $\mu\in (0,1/e]$. The random walk moves along open edges in $\mathbb{Z}^d$ (resp., open sites in $\mathcal{T}$) at rate $1$. For the critical regime $p=p_c$, we prove the following two results: on $\mathcal{T}$, the mean squared displacement of the random walk from $0$ to $t$ is at most $O(t\mu^{5/132-\epsilon})$ for any $\epsilon>0$; on $\mathbb{Z}^d$ with $d\geq 11$, the corresponding upper bound for the mean squared displacement is $O(t \mu^{1/2}\log(1/\mu))$. For the supercritical regime $p>p_c$, we prove that the mean squared displacement on $\mathbb{Z}^d$ is at least $ct$ for some $c=c(d)>0$ that does not depend on $\mu$., Comment: 26 pages, 1 figure

Let $G$ be a nonamenable transitive unimodular graph. In dynamical percolation, every edge in $G$ refreshes its status at rate $\mu>0$, and following the refresh, each edge is open independently with probability $p$. The random walk traverses $G$ only along open edges, moving at rate $1$. In the critical regime $p=p_c$, we prove that the speed of the random walk is at most $O(\sqrt{\mu \log(1/\mu)})$, provided that $\mu \le e^{-1}$. In the supercritical regime $p>p_c$, we prove that the speed on $G$ is of order 1 (uniformly in $\mu)$, while in the subcritical regime $p

Large language models (LLMs) have shown increasing capability in problem-solving and decision-making, largely based on the step-by-step chain-of-thought reasoning processes. However, evaluating these reasoning abilities has become increasingly challenging. Existing outcome-based benchmarks are beginning to saturate, becoming less effective in tracking meaningful progress. To address this, we present a process-based benchmark MR-Ben that demands a meta-reasoning skill, where LMs are asked to locate and analyse potential errors in automatically generated reasoning steps. Our meta-reasoning paradigm is especially suited for system-2 slow thinking, mirroring the human cognitive process of carefully examining assumptions, conditions, calculations, and logic to identify mistakes. MR-Ben comprises 5,975 questions curated by human experts across a wide range of subjects, including physics, chemistry, logic, coding, and more. Through our designed metrics for assessing meta-reasoning on this benchmark, we identify interesting limitations and weaknesses of current LLMs (open-source and closed-source models). For example, with models like the o1 series from OpenAI demonstrating strong performance by effectively scrutinizing the solution space, many other state-of-the-art models fall significantly behind on MR-Ben, exposing potential shortcomings in their training strategies and inference methodologies.

Integration of 2D materials in nanoelectromechanical systems (NEMS) marries the robustness of silicon-based materials with exceptional electrical controllability in 2D materials, drastically enhancing system performance which now is the key for many advanced applications in nanotechnology. Here, we experimentally demonstrate and theoretically analyze a powerful on-chip graphene integrated NEMS device consisting of a hybrid graphene/silicon-nitride membrane with metallic leads that enables an extremely large static and dynamic parameter regulation. When a static voltage is applied to the leads, the force induced by the thermal expansion difference between the leads and the membrane results in ultra-wide frequency tuning, deformation (post-buckling transition) and regulation of mechanical properties. Moreover, by injecting an alternating voltage to the leads, we can excite the resonator vibrating even far beyond its linear regime without a complex and space consuming actuation system. Our results prove that the device is a compact integrated system possessing mechanical robustness, high controllability, and fast response. It not only expands the limit of the application range of NEMS devices but also pushes multidimensional nanomechanical resonators into working in the nonlinear regime.

Federated Learning (FL) enables multiple devices to collaboratively train a shared model while ensuring data privacy. The selection of participating devices in each training round critically affects both the model performance and training efficiency, especially given the vast heterogeneity in training capabilities and data distribution across devices. To address these challenges, we introduce a novel device selection solution called FedRank, which is an end-to-end, ranking-based approach that is pre-trained by imitation learning against state-of-the-art analytical approaches. It not only considers data and system heterogeneity at runtime but also adaptively and efficiently chooses the most suitable clients for model training. Specifically, FedRank views client selection in FL as a ranking problem and employs a pairwise training strategy for the smart selection process. Additionally, an imitation learning-based approach is designed to counteract the cold-start issues often seen in state-of-the-art learning-based approaches. Experimental results reveal that \model~ boosts model accuracy by 5.2\% to 56.9\%, accelerates the training convergence up to $2.01 \times$ and saves the energy consumption up to $40.1\%$., Comment: Accepted by ICML 2024

Adapting Large Language Models (LLMs) to new tasks through fine-tuning has been made more efficient by the introduction of Parameter-Efficient Fine-Tuning (PEFT) techniques, such as LoRA. However, these methods often underperform compared to full fine-tuning, particularly in scenarios involving complex datasets. This issue becomes even more pronounced in complex domains, highlighting the need for improved PEFT approaches that can achieve better performance. Through a series of experiments, we have uncovered two critical insights that shed light on the training and parameter inefficiency of LoRA. Building on these insights, we have developed HydraLoRA, a LoRA framework with an asymmetric structure that eliminates the need for domain expertise. Our experiments demonstrate that HydraLoRA outperforms other PEFT approaches, even those that rely on domain knowledge during the training and inference phases., Comment: 19 pages, 7 figures

Modern speaker recognition system relies on abundant and balanced datasets for classification training. However, diverse defective datasets, such as partially-labelled, small-scale, and imbalanced datasets, are common in real-world applications. Previous works usually studied specific solutions for each scenario from the algorithm perspective. However, the root cause of these problems lies in dataset imperfections. To address these challenges with a unified solution, we propose the Voice Conversion Augmentation (VCA) strategy to obtain pseudo speech from the training set. Furthermore, to guarantee generation quality, we designed the VCA-NN~(nearest neighbours) strategy to select source speech from utterances that are close to the target speech in the representation space. Our experimental results on three created datasets demonstrated that VCA-NN effectively mitigates these dataset problems, which provides a new direction for handling the speaker recognition problems from the data aspect., Comment: 5 pages

In digital pathology, the multiple instance learning (MIL) strategy is widely used in the weakly supervised histopathology whole slide image (WSI) classification task where giga-pixel WSIs are only labeled at the slide level. However, existing attention-based MIL approaches often overlook contextual information and intrinsic spatial relationships between neighboring tissue tiles, while graph-based MIL frameworks have limited power to recognize the long-range dependencies. In this paper, we introduce the integrative graph-transformer framework that simultaneously captures the context-aware relational features and global WSI representations through a novel Graph Transformer Integration (GTI) block. Specifically, each GTI block consists of a Graph Convolutional Network (GCN) layer modeling neighboring relations at the local instance level and an efficient global attention model capturing comprehensive global information from extensive feature embeddings. Extensive experiments on three publicly available WSI datasets: TCGA-NSCLC, TCGA-RCC and BRIGHT, demonstrate the superiority of our approach over current state-of-the-art MIL methods, achieving an improvement of 1.0% to 2.6% in accuracy and 0.7%-1.6% in AUROC.

Large Language Models (LLMs) have succeeded remarkably in understanding long-form contents. However, exploring their capability for generating long-form contents, such as reports and articles, has been relatively unexplored and inadequately assessed by existing benchmarks. The prevalent evaluation methods, which predominantly rely on crowdsourcing, are recognized for their labor-intensive nature and lack of efficiency, whereas automated metrics, such as the ROUGE score, demonstrate discordance with human judgment criteria. In this paper, we propose ProxyQA, an innovative framework dedicated to assessing long-text generation. ProxyQA comprises in-depth human-curated meta-questions spanning various domains, each accompanied by specific proxy-questions with pre-annotated answers. LLMs are tasked to generate extensive content in response to these meta-questions, by engaging an evaluator and incorporating the generated texts as contextual background, ProxyQA assesses the generated content's quality through the evaluator's accuracy in addressing the proxy-questions. We examine multiple LLMs, emphasizing ProxyQA's demanding nature as a high-quality assessment tool. Human evaluation demonstrates that the proxy-question method is notably self-consistent and aligns closely with human evaluative standards. The dataset and leaderboard is available at \url{https://proxy-qa.com}., Comment: Accepted to ACL 2024 main conference

We demonstrate a compact, cost-effective snapshot spectral imaging system named Aperture Diffraction Imaging Spectrometer (ADIS), which consists only of an imaging lens with an ultra-thin orthogonal aperture mask and a mosaic filter sensor, requiring no additional physical footprint compared to common RGB cameras. Then we introduce a new optical design that each point in the object space is multiplexed to discrete encoding locations on the mosaic filter sensor by diffraction-based spatial-spectral projection engineering generated from the orthogonal mask. The orthogonal projection is uniformly accepted to obtain a weakly calibration-dependent data form to enhance modulation robustness. Meanwhile, the Cascade Shift-Shuffle Spectral Transformer (CSST) with strong perception of the diffraction degeneration is designed to solve a sparsity-constrained inverse problem, realizing the volume reconstruction from 2D measurements with Large amount of aliasing. Our system is evaluated by elaborating the imaging optical theory and reconstruction algorithm with demonstrating the experimental imaging under a single exposure. Ultimately, we achieve the sub-super-pixel spatial resolution and high spectral resolution imaging. The code will be available at: https://github.com/Krito-ex/CSST., Comment: accepted by International Conference on Computer Vision (ICCV) 2023

Ship orientation angle prediction (SOAP) with optical remote sensing images is an important image processing task, which often relies on deep convolutional neural networks (CNNs) to make accurate predictions. This paper proposes a novel framework to reduce the model sizes and computational costs of SOAP models without harming prediction accuracy. First, a new SOAP model called Mobile-SOAP is designed based on MobileNetV2, achieving state-of-the-art prediction accuracy. Four tiny SOAP models are also created by replacing the convolutional blocks in Mobile-SOAP with four small-scale networks, respectively. Then, to transfer knowledge from Mobile-SOAP to four lightweight models, we propose a novel knowledge distillation (KD) framework termed SOAP-KD consisting of a novel feature-based guidance loss and an optimized synthetic samples-based knowledge transfer mechanism. Lastly, extensive experiments on the FGSC-23 dataset confirm the superiority of Mobile-SOAP over existing models and also demonstrate the effectiveness of SOAP-KD in improving the prediction performance of four specially designed tiny models. Notably, by using SOAP-KD, the test mean absolute error of the ShuffleNetV2x1.0-based model is only 8% higher than that of Mobile-SOAP, but its number of parameters and multiply-accumulate operations (MACs) are respectively 61.6% and 60.8% less.

Shape transformation, a key mechanism for organismal survival and adaptation, has gained importance across fields as diverse as electronics and medicine. However, designing and controlling microscale shape-shifting materials remains a fundamental challenge in various actuation modalities. Here, we introduce SonoTransformer, an acoustically activated micromachine that delivers ultrafast transformability using preprogrammed soft hinges. These hinges concentrate energy through intensified oscillation and provide the necessary torque for transformation. We have created new machine designs to predetermine the folding state, enabling tailoring and milliseconds transformation. Additionally, we have shown selective transformation by adjusting acoustic power, realizing high degrees of control and functional versatility. Our findings open new research avenues in acoustics, physics, and soft matter, offering new design paradigms and development opportunities in robotics, metamaterials, adaptive optics, and microtechnology.

The problem of conditioning on the occupation field was investigated for the Brownian motion in 1998 independently by Aldous [4] and Warren and Yor [34] and recently for the loop soup at intensity $1/2$ by Werner [35], Sabot and Tarr\`es [30], and Lupu, Sabot and Tarr\`es [22]. We consider this problem in the case of the Brownian loop soup on the real line, and show that it is connected with a flow version of Jacobi processes, called Jacobi flow. We give a pathwise construction of this flow simultaneously for all parameters by means of a common Brownian motion, via the perturbed reflecting Brownian motion. The Jacobi flow is related to Fleming--Viot processes, as established by Bertoin and Le Gall [9] and Dawson and Li [11]. This relation allows us to interpret Perkins' disintegration theorem between Feller continuous state branching-processes and Fleming--Viot processes as a decomposition of Gaussian measures. Our approach gives a unified framework for the problems of disintegrating on the real line. The connection with Bass--Burdzy flows which was drawn in Warren [33] and Lupu, Sabot and Tarr\`es [23] is shown to be valid in the general case.

We consider a recursive system $(X_n)$ which was introduced by Collet et al. [10] as a spin glass model, and later by Derrida, Hakim, and Vannimenus [13] and by Derrida and Retaux [14] as a simplified hierarchical renormalization model. The system $(X_n)$ is expected to possess highly nontrivial universalities at or near criticality. In the nearly supercritical regime, Derrida and Retaux [14] conjectured that the free energy of the system decays exponentially with exponent $(p-p_c)^{-\frac12}$ as $p \downarrow p_c$. We study the nearly subcritical regime ($p \uparrow p_c$) and aim at a dual version of the Derrida-Retaux conjecture; our main result states that as $n \to \infty$, both $\E(X_n)$ and $\P(X_n\neq 0)$ decay exponentially with exponent $(p_c-p)^{\frac12 +o(1)}$, where $o(1) \to 0$ as $p \uparrow p_c$.

Neural network-based speaker recognition has achieved significant improvement in recent years. A robust speaker representation learns meaningful knowledge from both hard and easy samples in the training set to achieve good performance. However, noisy samples (i.e., with wrong labels) in the training set induce confusion and cause the network to learn the incorrect representation. In this paper, we propose a two-step audio-visual deep cleansing framework to eliminate the effect of noisy labels in speaker representation learning. This framework contains a coarse-grained cleansing step to search for the peculiar samples, followed by a fine-grained cleansing step to filter out the noisy labels. Our study starts from an efficient audio-visual speaker recognition system, which achieves a close to perfect equal-error-rate (EER) of 0.01\%, 0.07\% and 0.13\% on the Vox-O, E and H test sets. With the proposed multi-modal cleansing mechanism, four different speaker recognition networks achieve an average improvement of 5.9\%. Code has been made available at: \textcolor{magenta}{\url{https://github.com/TaoRuijie/AVCleanse}}., Comment: 5 pages, 3 figures

The success of scene graphs for visual scene understanding has brought attention to the benefits of abstracting a visual input (e.g., image) into a structured representation, where entities (people and objects) are nodes connected by edges specifying their relations. Building these representations, however, requires expensive manual annotation in the form of images paired with their scene graphs or frames. These formalisms remain limited in the nature of entities and relations they can capture. In this paper, we propose to leverage a widely-used meaning representation in the field of natural language processing, the Abstract Meaning Representation (AMR), to address these shortcomings. Compared to scene graphs, which largely emphasize spatial relationships, our visual AMR graphs are more linguistically informed, with a focus on higher-level semantic concepts extrapolated from visual input. Moreover, they allow us to generate meta-AMR graphs to unify information contained in multiple image descriptions under one representation. Through extensive experimentation and analysis, we demonstrate that we can re-purpose an existing text-to-AMR parser to parse images into AMRs. Our findings point to important future research directions for improved scene understanding., Comment: published in CoNLL 2022

We present a novel method for classification of Synthetic Aperture Radar (SAR) data by combining ideas from graph-based learning and neural network methods within an active learning framework. Graph-based methods in machine learning are based on a similarity graph constructed from the data. When the data consists of raw images composed of scenes, extraneous information can make the classification task more difficult. In recent years, neural network methods have been shown to provide a promising framework for extracting patterns from SAR images. These methods, however, require ample training data to avoid overfitting. At the same time, such training data are often unavailable for applications of interest, such as automatic target recognition (ATR) and SAR data. We use a Convolutional Neural Network Variational Autoencoder (CNNVAE) to embed SAR data into a feature space, and then construct a similarity graph from the embedded data and apply graph-based semi-supervised learning techniques. The CNNVAE feature embedding and graph construction requires no labeled data, which reduces overfitting and improves the generalization performance of graph learning at low label rates. Furthermore, the method easily incorporates a human-in-the-loop for active learning in the data-labeling process. We present promising results and compare them to other standard machine learning methods on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset for ATR with small amounts of labeled data.

Parking has been a painful problem for urban drivers. The parking pain exacerbates as more people tend to live in cities in the context of global urbanization. Thus, it is demanding to find a solution to mitigate d rivers' parking headaches. Many solutions tried to resolve the parking issue by predicting parking occupancy. Their focuses were on the accuracy of the theoretical side but lacked a standardized model to evaluate these proposals in practice. This paper develops a Driver Side and Traffic Based Evaluation Model (DSTBM), which provides a general evaluation scheme for different parking solutions. Two common parking detection methods, fixed sensing and mobile sensing are analyzed using DSTBM. The results indicate first, DSTBM examines different solutions from the driver's perspective and has no conflicts with other evaluation schemes; second, DSTBM confirms that fixed sensing performs better than mobile sensing in terms of prediction accuracy., Comment: 7 pages, 4 figures

While recent research on natural language inference has considerably benefited from large annotated datasets, the amount of inference-related knowledge (including commonsense) provided in the annotated data is still rather limited. There have been two lines of approaches that can be used to further address the limitation: (1) unsupervised pretraining can leverage knowledge in much larger unstructured text data; (2) structured (often human-curated) knowledge has started to be considered in neural-network-based models for NLI. An immediate question is whether these two approaches complement each other, or how to develop models that can bring together their advantages. In this paper, we propose models that leverage structured knowledge in different components of pre-trained models. Our results show that the proposed models perform better than previous BERT-based state-of-the-art models. Although our models are proposed for NLI, they can be easily extended to other sentence or sentence-pair classification problems.

Conversation disentanglement aims to separate intermingled messages into detached sessions, which is a fundamental task in understanding multi-party conversations. Existing work on conversation disentanglement relies heavily upon human-annotated datasets, which are expensive to obtain in practice. In this work, we explore to train a conversation disentanglement model without referencing any human annotations. Our method is built upon a deep co-training algorithm, which consists of two neural networks: a message-pair classifier and a session classifier. The former is responsible for retrieving local relations between two messages while the latter categorizes a message to a session by capturing context-aware information. Both networks are initialized respectively with pseudo data built from an unannotated corpus. During the deep co-training process, we use the session classifier as a reinforcement learning component to learn a session assigning policy by maximizing the local rewards given by the message-pair classifier. For the message-pair classifier, we enrich its training data by retrieving message pairs with high confidence from the disentangled sessions predicted by the session classifier. Experimental results on the large Movie Dialogue Dataset demonstrate that our proposed approach achieves competitive performance compared to the previous supervised methods. Further experiments show that the predicted disentangled conversations can promote the performance on the downstream task of multi-party response selection., Comment: Accepted to EMNLP 2021 main conference