arXiv subjects

Yu Qin

Publications and source records attributed to Yu Qin.

31 records · Page 2Linked to original sources

arXiv2023

Heterogeneous Knowledge Fusion: A Novel Approach for Personalized Recommendation via LLM

The analysis and mining of user heterogeneous behavior are of paramount importance in recommendation systems. However, the conventional approach of incorporating various types of heterogeneous behavior into recommendation models leads to feature sparsity and knowledge fragmentation issues. To address this challenge, we propose a novel approach for personalized recommendation via Large Language Model (LLM), by extracting and fusing heterogeneous knowledge from user heterogeneous behavior information. In addition, by combining heterogeneous knowledge and recommendation tasks, instruction tuning is performed on LLM for personalized recommendations. The experimental results demonstrate that our method can effectively integrate user heterogeneous behavior and significantly improve recommendation performance.

cs.IR↗

arXiv2023

Dynamic Grouping for Climate Change Negotiation: Facilitating Cooperation and Balancing Interests through Effective Strategies

The current framework for climate change negotiation models presents several limitations that warrant further research and development. In this track, we discuss mainly two key areas for improvement, focusing on the geographical impacts and utility framework. In the aspects of geographical impacts, We explore five critical aspects: (1) the shift from local to global impact, (2) variability in climate change effects across regions, (3) heterogeneity in geographical location and political structures, and (4) collaborations between adjacent nations, (5) the importance of including historical and cultural factors influencing climate negotiations. Furthermore, we emphasize the need to refine the utility and rewards framework to reduce the homogeneity and the level of overestimating the climate mitigation by integrating the positive effects of saving rates into the reward function and heterogeneity among all regions. By addressing these limitations, we hope to enhance the accuracy and effectiveness of climate change negotiation models, enabling policymakers and stakeholders to devise targeted and appropriate strategies to tackle climate change at both regional and global levels.

cs.CY↗

arXiv2023

Dynamic Grouping for Climate Change Negotiation: Facilitating Cooperation and Balancing Interests through Effective Strategies

In this paper, we propose a dynamic grouping negotiation model for climate mitigation based on real-world business and political negotiation protocols. Within the AI4GCC competition framework, we develop a three-stage process: group formation and updates, intra-group negotiation, and inter-group negotiation. Our model promotes efficient and effective cooperation between various stakeholders to achieve global climate change objectives. By implementing a group-forming method and group updating strategy, we address the complexities and imbalances in multi-region climate negotiations. Intra-group negotiations ensure that all members contribute to mitigation efforts, while inter-group negotiations use the proposal-evaluation framework to set mitigation and savings rates. We demonstrate our negotiation model within the RICE-N framework, illustrating a promising approach for facilitating international cooperation on climate change mitigation.

cs.CY↗

arXiv2022

PANet: Perspective-Aware Network with Dynamic Receptive Fields and Self-Distilling Supervision for Crowd Counting

Crowd counting aims to learn the crowd density distributions and estimate the number of objects (e.g. persons) in images. The perspective effect, which significantly influences the distribution of data points, plays an important role in crowd counting. In this paper, we propose a novel perspective-aware approach called PANet to address the perspective problem. Based on the observation that the size of the objects varies greatly in one image due to the perspective effect, we propose the dynamic receptive fields (DRF) framework. The framework is able to adjust the receptive field by the dilated convolution parameters according to the input image, which helps the model to extract more discriminative features for each local region. Different from most previous works which use Gaussian kernels to generate the density map as the supervised information, we propose the self-distilling supervision (SDS) training method. The ground-truth density maps are refined from the first training stage and the perspective information is distilled to the model in the second stage. The experimental results on ShanghaiTech Part_A and Part_B, UCF_QNRF, and UCF_CC_50 datasets demonstrate that our proposed PANet outperforms the state-of-the-art methods by a large margin.

cs.CV↗

arXiv2022

Mesoscopic Collective Activity in Excitatory Neural Fields: Governing Equations

In this study we derive the governing equations for mesoscopic collective activity in the cortex, starting from the generic Hodgkin-Huxley equations for microscopic cell dynamics. For simplicity, and to maintain focus on the essential elements of the derivation, the discussion is confined to excitatory neural fields. The fundamental assumption of the procedure is that mesoscale processes are macroscopic with respect to cell-scale activity, and emerge as the average behavior of a large population of cells. Because of their duration, action-potential details are assumed not observable at mesoscale; the essential mesoscopic function of action potentials is to redistribute energy in the neural field. The Hodgkin-Huxley dynamical model is first reduced to a set of equations that describe subthreshold dynamics. An ensemble average over a cell population then produces a closed system of equations involving two mesoscopic state variables: the density of kinetic energy J, carried by sodium ionic currents, and the excitability H of the neural field, which could be described as the average state of gating variable h. The resulting model is represented as essentially a subthreshold process; and the dynamical role of the firing rate is naturally reassessed as describing energy transfers. The linear properties of the equations are consistent with expectations for the dynamics of excitatory neural fields: the system supports oscillations of progressive waves, with shorter waves typically having higher frequencies, propagating slower, and decaying faster. Extending the derivation to include more complex cell dynamics (e.g., including other ionic channels, e.g., calcium channels) and multiple-type, excitatory-inhibitory, neural fields is straightforward, and will be presented elsewhere.

q-bio.QM↗

arXiv2022

Mesoscopic Collective Activity in Excitatory Neural Fields: Cross-frequency Coupling

In the brain, cross-frequency coupling has been hypothesized to result from the activity of specialized microcircuits. For example, theta-gamma coupling is assumed to be generated by specialized cell pairs (PING and ING mechanisms), or special cells (e.g., fast bursting neurons). However, this implies that the generating mechanisms is uniquely specific to the brain. In fact, cross-scale coupling is a phenomenon encountered in the physics of all large, multi-scale systems: phase and amplitude correlations between components of different scales arise as a result of nonlinear interaction. Because the brain is a multi-scale system too, a similar mechanism must be active in the brain. Here, we represent brain activity as a superposition of nonlinearly interacting patterns of spatio-temporal activity (collective activity), supported by populations of neurons. Cross-frequency coupling is a direct consequence of the nonlinear interactions, and does not require specialized cells or cell pairs. It is therefore universal, and must be active in neural fields of any composition. To emphasize this, we demonstrate the phenomenon in excitatory fields. While there is no doubt that specialized cells play a role in theta-gamma coupling, our results suggest that the coupling mechanism is at the same time simpler and richer: simpler because it involves the universal principle of nonlinearity; richer, because nonlinearity of collective activity is likely modulated by specialized-cell populations in ways to be yet understood.

q-bio.NC↗

arXiv2021

Deconfounded Visual Grounding

We focus on the confounding bias between language and location in the visual grounding pipeline, where we find that the bias is the major visual reasoning bottleneck. For example, the grounding process is usually a trivial language-location association without visual reasoning, e.g., grounding any language query containing sheep to the nearly central regions, due to that most queries about sheep have ground-truth locations at the image center. First, we frame the visual grounding pipeline into a causal graph, which shows the causalities among image, query, target location and underlying confounder. Through the causal graph, we know how to break the grounding bottleneck: deconfounded visual grounding. Second, to tackle the challenge that the confounder is unobserved in general, we propose a confounder-agnostic approach called: Referring Expression Deconfounder (RED), to remove the confounding bias. Third, we implement RED as a simple language attention, which can be applied in any grounding method. On popular benchmarks, RED improves various state-of-the-art grounding methods by a significant margin. Code will soon be available at: https://github.com/JianqiangH/Deconfounded_VG.

cs.CV↗

arXiv2021

A Domain-Oblivious Approach for Learning Concise Representations of Filtered Topological Spaces for Clustering

Persistence diagrams have been widely used to quantify the underlying features of filtered topological spaces in data visualization. In many applications, computing distances between diagrams is essential; however, computing these distances has been challenging due to the computational cost. In this paper, we propose a persistence diagram hashing framework that learns a binary code representation of persistence diagrams, which allows for fast computation of distances. This framework is built upon a generative adversarial network (GAN) with a diagram distance loss function to steer the learning process. Instead of using standard representations, we hash diagrams into binary codes, which have natural advantages in large-scale tasks. The training of this model is domain-oblivious in that it can be computed purely from synthetic, randomly created diagrams. As a consequence, our proposed method is directly applicable to various datasets without the need for retraining the model. These binary codes, when compared using fast Hamming distance, better maintain topological similarity properties between datasets than other vectorized representations. To evaluate this method, we apply our framework to the problem of diagram clustering and we compare the quality and performance of our approach to the state-of-the-art. In addition, we show the scalability of our approach on a dataset with 10k persistence diagrams, which is not possible with current techniques. Moreover, our experimental results demonstrate that our method is significantly faster with the potential of less memory usage, while retaining comparable or better quality comparisons.

cs.CG↗

arXiv2019

Knowledge Transfer between Datasets for Learning-based Tissue Microstructure Estimation

Learning-based approaches, especially those based on deep networks, have enabled high-quality estimation of tissue microstructure from low-quality diffusion magnetic resonance imaging (dMRI) scans, which are acquired with a limited number of diffusion gradients and a relatively poor spatial resolution. These learning-based approaches to tissue microstructure estimation require acquisitions of training dMRI scans with high-quality diffusion signals, which are densely sampled in the q-space and have a high spatial resolution. However, the acquisition of training scans may not be available for all datasets. Therefore, we explore knowledge transfer between different dMRI datasets so that learning-based tissue microstructure estimation can be applied for datasets where training scans are not acquired. Specifically, for a target dataset of interest, where only low-quality diffusion signals are acquired without training scans, we exploit the information in a source dMRI dataset acquired with high-quality diffusion signals. We interpolate the diffusion signals in the source dataset in the q-space using a dictionary-based signal representation, so that the interpolated signals match the acquisition scheme of the target dataset. Then, the interpolated signals are used together with the high-quality tissue microstructure computed from the source dataset to train deep networks that perform tissue microstructure estimation for the target dataset. Experiments were performed on brain dMRI scans with low-quality diffusion signals, where the benefit of the proposed strategy is demonstrated.

eess.IV↗

arXiv2019

Attend More Times for Image Captioning

Most attention-based image captioning models attend to the image once per word. However, attending once per word is rigid and is easy to miss some information. Attending more times can adjust the attention position, find the missing information back and avoid generating the wrong word. In this paper, we show that attending more times per word can gain improvements in the image captioning task, without increasing the number of parameters. We propose a flexible two-LSTM merge model to make it convenient to encode more attentions than words. Our captioning model uses two LSTMs to encode the word sequence and the attention sequence respectively. The information of the two LSTMs and the image feature are combined to predict the next word. Experiments on the MSCOCO caption dataset show that our method outperforms the state-of-the-art. Using bottom up features and self-critical training method, our method gets BLEU-4, METEOR, ROUGE-L, CIDEr and SPICE scores of 0.381, 0.283, 0.580, 1.261 and 0.220 on the Karpathy test split.

cs.CV↗

arXiv2017

Quantum Inspired Security on a Mobile Phone

The widespread use of mobile electronic devices increases the complexities of mobile security. This paper aims to provide a secure communication environment for smartphone users. Some research proves that the one-time pad is one of the securest encryption methods, and the key distribution problem can be solved by using the QKD (quantum key distribution). The objective of this project is to design an Android APP (application) to exchange several random keys between mobile phones. Inspired by QKD, the developed APP uses the quick response (QR) code as a carrier to dispatch large amounts of one-time keys. After evaluating the performance of APP, it allows the mobile phone to capture and decode 1800 bytes of random data in 600ms. The continuous scanning mode of APP is designed to improve the overall transmission performance and user experience, and the maximum transmission rate of this mode is around 2200 bytes/s. The omnidirectional readability and error correction capability of QR code gives it better real-life application, and the features of adequate storage capacity and quick response optimize overall transmission efficiency. The security of this APP is guaranteed since QR code is exchanged face-to-face, eliminating the risk of being eavesdropped. Also, the id of QR code is the only message that would be transmitted through the whole communication. The experimental results show this project can achieve superior transmission performance, and the correlation between the transmission rate of the system and several parameters, such as the QR code size, has been analyzed. In addition, some existing technologies and the main findings in the context of the project are summarized and critically compared in detail.

cs.CR↗

arXiv2016

Computational Soundness Results for Stateful Applied pi Calculus

In recent years, many researches have been done to establish symbolic models of stateful protocols. Two works among them, the SAPIC tool and StatVerif tool, provide a high-level specification language and an automated analysis. Their language, the stateful applied π-calculus, is extended from the applied π-calculus by defining explicit state constructs. Symbolic abstractions of cryptography used in it make the analysis amenable to automation. However, this might overlook the attacks based on the algebraic properties of the cryptographic algorithms. In our paper, we establish the computational soundness results for stateful applied π-calculus used in SAPIC tool and StatVerif tool. In our approach, we build our results on the CoSP framework. For SAPIC, we embed the non-monotonic protocol states into the CoSP protocols, and prove that the resulting CoSP protocols are efficient. Through the embedding, we provide the computational soundness result for SAPIC (by Theorem 1). For StatVerif, we encode the StatVerif process into a subset of SAPIC process, and obtain the computational soundness result for StatVerif (by Theorem 2). Our encoding shows the differences between the semantics of the two languages. Our work inherits the modularity of CoSP, which allows for easily extending the proofs to specific cryptographic primitives. Thus we establish a computationally sound automated verification result for the input languages of SAPIC and StatVerif that use public-key encryption and signatures (by Theorem 3).

cs.CR↗

arXiv2013

Diagnosis of the Wavelength Stability of a Mid-Infrared Free-Electron Laser

Wavelength stability of free-electron lasers (FELs) is one of the important parameters for various applications. In this paper we describe two different methods to diagnose the wavelength stability of a mid-infrared (MIR) FEL. The first one is based on autocorrelation which is usually used to measure the pulse duration, and the second one is based on frequency upconversion through sum-frequency mixing (SFM).

physics.optics↗