Searcharxiv⌕ Search

arXiv subjects

Lin Chen

Publications and source records attributed to Lin Chen.

At least 271 records · Page 15Linked to original sources
arXiv2023

ShareGPT4V: Improving Large Multi-Modal Models with Better Captions

In the realm of large multi-modal models (LMMs), efficient modality alignment is crucial yet often constrained by the scarcity of high-quality image-text data. To address this bottleneck, we introduce the ShareGPT4V dataset, a pioneering large-scale resource featuring 1.2 million highly descriptive captions, which surpasses existing datasets in diversity and information content, covering world knowledge, object properties, spatial relationships, and aesthetic evaluations. Specifically, ShareGPT4V originates from a curated 100K high-quality captions collected from advanced GPT4-Vision and has been expanded to 1.2M with a superb caption model trained on this subset. ShareGPT4V first demonstrates its effectiveness for the Supervised Fine-Tuning (SFT) phase, by substituting an equivalent quantity of detailed captions in existing SFT datasets with a subset of our high-quality captions, significantly enhancing the LMMs like LLaVA-7B, LLaVA-1.5-13B, and Qwen-VL-Chat-7B on the MME and MMBench benchmarks, with respective gains of 222.8/22.0/22.3 and 2.7/1.3/1.5. We further incorporate ShareGPT4V data into both the pre-training and SFT phases, obtaining ShareGPT4V-7B, a superior LMM based on a simple architecture that has remarkable performance across a majority of the multi-modal benchmarks. This project is available at https://ShareGPT4V.github.io to serve as a pivotal resource for advancing the LMMs community.

cs.CV↗
arXiv2023

The First LHAASO Catalog of Gamma-Ray Sources

We present the first catalog of very-high energy and ultra-high energy gamma-ray sources detected by the Large High Altitude Air Shower Observatory (LHAASO). The catalog was compiled using 508 days of data collected by the Water Cherenkov Detector Array (WCDA) from March 2021 to September 2022 and 933 days of data recorded by the Kilometer Squared Array (KM2A) from January 2020 to September 2022. This catalog represents the main result from the most sensitive large coverage gamma-ray survey of the sky above 1 TeV, covering declination from $-$20$^{\circ}$ to 80$^{\circ}$. In total, the catalog contains 90 sources with an extended size smaller than $2^\circ$ and a significance of detection at $> 5σ$. Based on our source association criteria, 32 new TeV sources are proposed in this study. Among the 90 sources, 43 sources are detected with ultra-high energy ($E > 100$ TeV) emission at $> 4σ$ significance level. We provide the position, extension, and spectral characteristics of all the sources in this catalog.

astro-ph.HE↗
arXiv2023

Very high energy gamma-ray emission beyond 10 TeV from GRB 221009A

The highest energy gamma-rays from gamma-ray bursts (GRBs) have important implications for their radiation mechanism. Here we report for the first time the detection of gamma-rays up to 13 TeV from the brightest GRB 221009A by the Large High Altitude Air-shower Observatory (LHAASO). The LHAASO-KM2A detector registered more than 140 gamma-rays with energies above 3 TeV during 230$-$900s after the trigger. The intrinsic energy spectrum of gamma-rays can be described by a power-law after correcting for extragalactic background light (EBL) absorption. Such a hard spectrum challenges the synchrotron self-Compton (SSC) scenario of relativistic electrons for the afterglow emission above several TeV. Observations of gamma-rays up to 13 TeV from a source with a measured redshift of z=0.151 hints more transparency in intergalactic space than previously expected. Alternatively, one may invoke new physics such as Lorentz Invariance Violation (LIV) or an axion origin of very high energy (VHE) signals.

astro-ph.HE↗
arXiv2023

CurriculumLoc: Enhancing Cross-Domain Geolocalization through Multi-Stage Refinement

Visual geolocalization is a cost-effective and scalable task that involves matching one or more query images, taken at some unknown location, to a set of geo-tagged reference images. Existing methods, devoted to semantic features representation, evolving towards robustness to a wide variety between query and reference, including illumination and viewpoint changes, as well as scale and seasonal variations. However, practical visual geolocalization approaches need to be robust in appearance changing and extreme viewpoint variation conditions, while providing accurate global location estimates. Therefore, inspired by curriculum design, human learn general knowledge first and then delve into professional expertise. We first recognize semantic scene and then measure geometric structure. Our approach, termed CurriculumLoc, involves a delicate design of multi-stage refinement pipeline and a novel keypoint detection and description with global semantic awareness and local geometric verification. We rerank candidates and solve a particular cross-domain perspective-n-point (PnP) problem based on these keypoints and corresponding descriptors, position refinement occurs incrementally. The extensive experimental results on our collected dataset, TerraTrack and a benchmark dataset, ALTO, demonstrate that our approach results in the aforementioned desirable characteristics of a practical visual geolocalization solution. Additionally, we achieve new high recall@1 scores of 62.6% and 94.5% on ALTO, with two different distances metrics, respectively. Dataset, code and trained models are publicly available on https://github.com/npupilab/CurriculumLoc.

cs.CV↗
arXiv2023

Greedy PIG: Adaptive Integrated Gradients

Deep learning has become the standard approach for most machine learning tasks. While its impact is undeniable, interpreting the predictions of deep learning models from a human perspective remains a challenge. In contrast to model training, model interpretability is harder to quantify and pose as an explicit optimization problem. Inspired by the AUC softmax information curve (AUC SIC) metric for evaluating feature attribution methods, we propose a unified discrete optimization framework for feature attribution and feature selection based on subset selection. This leads to a natural adaptive generalization of the path integrated gradients (PIG) method for feature attribution, which we call Greedy PIG. We demonstrate the success of Greedy PIG on a wide variety of tasks, including image feature attribution, graph compression/explanation, and post-hoc feature selection on tabular data. Our results show that introducing adaptivity is a powerful and versatile method for making attribution methods more powerful.

cs.LG↗
arXiv2023

Does or did the supernova remnant Cassiopeia A operate as a PeVatron?

For decades, supernova remnants (SNRs) have been considered the prime sources of Galactic Cosmic rays (CRs). But whether SNRs can accelerate CR protons to PeV energies and thus dominate CR flux up to the knee is currently under intensive theoretical and phenomenological debate. The direct test of the ability of SNRs to operate as CR PeVatrons can be provided by ultrahigh-energy (UHE; $E_γ\geq 100$~TeV) $γ$-rays. In this context, the historical SNR Cassiopeia A (Cas A) is considered one of the most promising target for UHE observations. This paper presents the observation of Cas A and its vicinity by the LHAASO KM2A detector. The exceptional sensitivity of LHAASO KM2A in the UHE band, combined with the young age of Cas A, enabled us to derive stringent model-independent limits on the energy budget of UHE protons and nuclei accelerated by Cas A at any epoch after the explosion. The results challenge the prevailing paradigm that Cas A-type SNRs are major suppliers of PeV CRs in the Milky Way.

astro-ph.HE↗
arXiv2023

Inertia of partial transpose of positive semidefinite matrices

We show that the partial transpose of $9\times 9$ positive semidefinite matrices do not have inertia (4,1,4) and (3,2,4). It solves an open problem in "LINEAR AND MULTILINEAR ALGEBRA, Changchun Feng et al, 2022". We apply our results to construct some inertia, as well as present the list of all possible inertia of partial transpose of $12\times 12$ positive semidefinite matrices.

quant-ph↗
arXiv2023

It's an Alignment, Not a Trade-off: Revisiting Bias and Variance in Deep Models

Classical wisdom in machine learning holds that the generalization error can be decomposed into bias and variance, and these two terms exhibit a \emph{trade-off}. However, in this paper, we show that for an ensemble of deep learning based classification models, bias and variance are \emph{aligned} at a sample level, where squared bias is approximately \emph{equal} to variance for correctly classified sample points. We present empirical evidence confirming this phenomenon in a variety of deep learning models and datasets. Moreover, we study this phenomenon from two theoretical perspectives: calibration and neural collapse. We first show theoretically that under the assumption that the models are well calibrated, we can observe the bias-variance alignment. Second, starting from the picture provided by the neural collapse theory, we show an approximate correlation between bias and variance.

cs.LG↗
arXiv2023

Nonlocality via multiqubit orthogonal product bases

We investigate the quantum nonlocality via the discrimination on two, three and four-qubit orthogonal product bases (OPBs). We show that every two-qubit, and some three and four-qubit OPBs can be locally distinguished. It turns out that the remaining three and four-qubit OPBs cannot be locally distinguished, and thus they show the quantum nonlocality without entanglement. We also distinguish them by merging some systems using quantum teleportation with assisted Bell states.

quant-ph↗
arXiv2023

Mutually-orthogonal unitary and orthogonal matrices

We introduce the concept of n-OU and n-OO matrix sets, a collection of n mutually-orthogonal unitary and real orthogonal matrices under Hilbert-Schmidt inner product. We give a detailed characterization of order-three n-OO matrix sets under orthogonal equivalence. As an application in quantum information theory, we show that the minimum and maximum numbers of an unextendible maximally entangled bases within a real two-qutrit system are three and four, respectively. Further, we propose a new matrix decomposition approach, defining an n-OU (resp. n-OO) decomposition for a matrix as a linear combination of n matrices from an n-OU (resp. n-OO) matrix set. We show that any order-d matrix has a d-OU decomposition. As a contrast, we provide criteria for an order-three real matrix to possess an n-OO decomposition.

quant-ph↗
arXiv2023

Chained-DP: Can We Recycle Privacy Budget?

Privacy-preserving vector mean estimation is a crucial primitive in federated analytics. Existing practices usually resort to Local Differentiated Privacy (LDP) mechanisms that inject random noise into users' vectors when communicating with users and the central server. Due to the privacy-utility trade-off, the privacy budget has been widely recognized as the bottleneck resource that requires well-provisioning. In this paper, we explore the possibility of privacy budget recycling and propose a novel Chained-DP framework enabling users to carry out data aggregation sequentially to recycle the privacy budget. We establish a sequential game to model the user interactions in our framework. We theoretically show the mathematical nature of the sequential game, solve its Nash Equilibrium, and design an incentive mechanism with provable economic properties. We further derive a differentially privacy-guaranteed protocol to alleviate potential privacy collusion attacks to avoid holistic exposure. Our numerical simulation validates the effectiveness of Chained-DP, showing that it can significantly save privacy budget and lower estimation error compared to the traditional LDP mechanism.

cs.CR↗
arXiv2023

Roulette: A Semantic Privacy-Preserving Device-Edge Collaborative Inference Framework for Deep Learning Classification Tasks

Deep learning classifiers are crucial in the age of artificial intelligence. The device-edge-based collaborative inference has been widely adopted as an efficient framework for promoting its applications in IoT and 5G/6G networks. However, it suffers from accuracy degradation under non-i.i.d. data distribution and privacy disclosure. For accuracy degradation, direct use of transfer learning and split learning is high cost and privacy issues remain. For privacy disclosure, cryptography-based approaches lead to a huge overhead. Other lightweight methods assume that the ground truth is non-sensitive and can be exposed. But for many applications, the ground truth is the user's crucial privacy-sensitive information. In this paper, we propose a framework of Roulette, which is a task-oriented semantic privacy-preserving collaborative inference framework for deep learning classifiers. More than input data, we treat the ground truth of the data as private information. We develop a novel paradigm of split learning where the back-end DNN is frozen and the front-end DNN is retrained to be both a feature extractor and an encryptor. Moreover, we provide a differential privacy guarantee and analyze the hardness of ground truth inference attacks. To validate the proposed Roulette, we conduct extensive performance evaluations using realistic datasets, which demonstrate that Roulette can effectively defend against various attacks and meanwhile achieve good model accuracy. In a situation where the non-i.i.d. is very severe, Roulette improves the inference accuracy by 21\% averaged over benchmarks, while making the accuracy of discrimination attacks almost equivalent to random guessing.

cs.LG↗
arXiv2023

The additivity of states uniquely determined by marginals

The pure states that can be uniquely determined among all (UDA) states by their marginals are essential to efficient quantum state tomography. We generalize the UDA states from the context of pure states to that of arbitrary (no matter pure or mixed) states, motivated by the efficient state tomography of low-rank states. We call the \emph{additivity} of $k$-UDA states for three different composite ways of tensor product, if the composite state of two $k$-UDA states is still uniquely determined by the $k$-partite marginals for the corresponding type of tensor product. We show that the additivity holds if one of the two initial states is pure, and present the conditions under which the additivity holds for two mixed UDA states. One of the three composite ways of tensor product is also adopted to construct genuinely multipartite entangled (GME) states. Therefore, it is effective to construct multipartite $k$-UDA state with genuine entanglement by uniting the additivity of $k$-UDA states and the construction of GME states.

quant-ph↗
arXiv2023

Measurement of ultra-high-energy diffuse gamma-ray emission of the Galactic plane from 10 TeV to 1 PeV with LHAASO-KM2A

The diffuse Galactic $γ$-ray emission, mainly produced via interactions between cosmic rays and the interstellar medium and/or radiation field, is a very important probe of the distribution, propagation, and interaction of cosmic rays in the Milky Way. In this work we report the measurements of diffuse $γ$-rays from the Galactic plane between 10 TeV and 1 PeV energies, with the square kilometer array of the Large High Altitude Air Shower Observatory (LHAASO). Diffuse emissions from the inner ($15^{\circ} 10$~TeV). The energy spectrum in the inner Galaxy regions can be described by a power-law function with an index of $-2.99\pm0.04$, which is different from the curved spectrum as expected from hadronic interactions between locally measured cosmic rays and the line-of-sight integrated gas content. Furthermore, the measured flux is higher by a factor of $\sim3$ than the prediction. A similar spectrum with an index of $-2.99\pm0.07$ is found in the outer Galaxy region, and the absolute flux for $10\lesssim E\lesssim60$ TeV is again higher than the prediction for hadronic cosmic ray interactions. The latitude distributions of the diffuse emission are consistent with the gas distribution, while the longitude distributions show clear deviation from the gas distribution. The LHAASO measurements imply that either additional emission sources exist or cosmic ray intensities have spatial variations.

astro-ph.HE↗
arXiv2023

Efficient Commercial Bank Customer Credit Risk Assessment Based on LightGBM and Feature Engineering

Effective control of credit risk is a key link in the steady operation of commercial banks. This paper is mainly based on the customer information dataset of a foreign commercial bank in Kaggle, and we use LightGBM algorithm to build a classifier to classify customers, to help the bank judge the possibility of customer credit default. This paper mainly deals with characteristic engineering, such as missing value processing, coding, imbalanced samples, etc., which greatly improves the machine learning effect. The main innovation of this paper is to construct new feature attributes on the basis of the original dataset so that the accuracy of the classifier reaches 0.734, and the AUC reaches 0.772, which is more than many classifiers based on the same dataset. The model can provide some reference for commercial banks' credit granting, and also provide some feature processing ideas for other similar studies.

cs.LG↗
arXiv2023

Improved test-retest reliability of $\textit{R}_2^*$ and susceptibility quantification using multi-shot multi echo 3D EPI

This study aimed to evaluate the potential of 3D echo-planar imaging (EPI) for improving the reliability of $T_2^*$-weighted ($T_2^*w$) data and quantification of $\textit{R}_2^*$ decay rate and susceptibility ($χ$) compared to conventional gradient echo (GRE)-based acquisition. Eight healthy subjects in a wide age range were recruited. Each subject received repeated scans for both GRE and EPI acquisitions with an isotropic 1 mm resolution at 3 T. Maps of $\textit{R}_2^*$ and $χ$ were quantified and compared using their inter-scan difference to evaluate the test-retest reliability. Inter-protocol differences of $\textit{R}_2^*$ and $χ$ between GRE and EPI were also measured voxel by voxel and in selected ROIs to test the consistency between the two acquisition methods. The quantifications of $\textit{R}_2^*$ and $χ$ using EPI protocols showed increased test-retest reliability with higher EPI factors up to 5 as performed in the experiment and were consistent with those based on GRE. This result suggested multi-shot multi-echo 3D EPI can be a useful alternative acquisition method for $T_2^*w$ MRI and quantification of $\textit{R}_2^*$ and $χ$ with reduced scan time, improved test-retest reliability and similar accuracy compared to commonly used 3D GRE.

physics.med-ph↗
arXiv2023

An End-to-End Framework of Road User Detection, Tracking, and Prediction from Monocular Images

Perception that involves multi-object detection and tracking, and trajectory prediction are two major tasks of autonomous driving. However, they are currently mostly studied separately, which results in most trajectory prediction modules being developed based on ground truth trajectories without taking into account that trajectories extracted from the detection and tracking modules in real-world scenarios are noisy. These noisy trajectories can have a significant impact on the performance of the trajectory predictor and can lead to serious prediction errors. In this paper, we build an end-to-end framework for detection, tracking, and trajectory prediction called ODTP (Online Detection, Tracking and Prediction). It adopts the state-of-the-art online multi-object tracking model, QD-3DT, for perception and trains the trajectory predictor, DCENet++, directly based on the detection results without purely relying on ground truth trajectories. We evaluate the performance of ODTP on the widely used nuScenes dataset for autonomous driving. Extensive experiments show that ODPT achieves high performance end-to-end trajectory prediction. DCENet++, with the enhanced dynamic maps, predicts more accurate trajectories than its base model. It is also more robust when compared with other generative and deterministic trajectory prediction models trained on noisy detection results.

cs.CV↗
arXiv2023

Imbalanced Large Graph Learning Framework for FPGA Logic Elements Packing Prediction

Packing is a required step in a typical FPGA CAD flow. It has high impacts to the performance of FPGA placement and routing. Early prediction of packing results can guide design optimization and expedite design closure. In this work, we propose an imbalanced large graph learning framework, ImLG, for prediction of whether logic elements will be packed after placement. Specifically, we propose dedicated feature extraction and feature aggregation methods to enhance the node representation learning of circuit graphs. With imbalanced distribution of packed and unpacked logic elements, we further propose techniques such as graph oversampling and mini-batch training for this imbalanced learning task in large circuit graphs. Experimental results demonstrate that our framework can improve the F1 score by 42.82% compared to the most recent Gaussian-based prediction method. Physical design results show that the proposed method can assist the placer in improving routed wirelength by 0.93% and SLICE occupation by 0.89%.

cs.LG↗