Searcharxiv⌕ Search

arXiv subjects

Chen Sun

Publications and source records attributed to Chen Sun.

At least 145 records · Page 8Linked to original sources
arXiv2023

2nd Place Winning Solution for the CVPR2023 Visual Anomaly and Novelty Detection Challenge: Multimodal Prompting for Data-centric Anomaly Detection

This technical report introduces the winning solution of the team Segment Any Anomaly for the CVPR2023 Visual Anomaly and Novelty Detection (VAND) challenge. Going beyond uni-modal prompt, e.g., language prompt, we present a novel framework, i.e., Segment Any Anomaly + (SAA$+$), for zero-shot anomaly segmentation with multi-modal prompts for the regularization of cascaded modern foundation models. Inspired by the great zero-shot generalization ability of foundation models like Segment Anything, we first explore their assembly (SAA) to leverage diverse multi-modal prior knowledge for anomaly localization. Subsequently, we further introduce multimodal prompts (SAA$+$) derived from domain expert knowledge and target image context to enable the non-parameter adaptation of foundation models to anomaly segmentation. The proposed SAA$+$ model achieves state-of-the-art performance on several anomaly segmentation benchmarks, including VisA and MVTec-AD, in the zero-shot setting. We will release the code of our winning solution for the CVPR2023 VAN.

cs.CV↗
arXiv2023

NLOS Dies Twice: Challenges and Solutions of V2X for Cooperative Perception

Multi-agent multi-lidar sensor fusion between connected vehicles for cooperative perception has recently been recognized as the best technique for minimizing the blind zone of individual vehicular perception systems and further enhancing the overall safety of autonomous driving systems. This technique relies heavily on the reliability and availability of vehicle-to-everything (V2X) communication. In practical sensor fusion application scenarios, the non-line-of-sight (NLOS) issue causes blind zones for not only the perception system but also V2X direct communication. To counteract underlying communication issues, we introduce an abstract perception matrix matching method for quick sensor fusion matching procedures and mobility-height hybrid relay determination procedures, proactively improving the efficiency and performance of V2X communication to serve the upper layer application fusion requirements. To demonstrate the effectiveness of our solution, we design a new simulation framework to consider autonomous driving, sensor fusion and V2X communication in general, paving the way for end-to-end performance evaluation and further solution derivation.

eess.SY↗
arXiv2023

Federated Learning over a Wireless Network: Distributed User Selection through Random Access

User selection has become crucial for decreasing the communication costs of federated learning (FL) over wireless networks. However, centralized user selection causes additional system complexity. This study proposes a network intrinsic approach of distributed user selection that leverages the radio resource competition mechanism in random access. Taking the carrier sensing multiple access (CSMA) mechanism as an example of random access, we manipulate the contention window (CW) size to prioritize certain users for obtaining radio resources in each round of training. Training data bias is used as a target scenario for FL with user selection. Prioritization is based on the distance between the newly trained local model and the global model of the previous round. To avoid excessive contribution by certain users, a counting mechanism is used to ensure fairness. Simulations with various datasets demonstrate that this method can rapidly achieve convergence similar to that of the centralized user selection approach.

cs.LG↗
arXiv2023

How can objects help action recognition?

Current state-of-the-art video models process a video clip as a long sequence of spatio-temporal tokens. However, they do not explicitly model objects, their interactions across the video, and instead process all the tokens in the video. In this paper, we investigate how we can use knowledge of objects to design better video models, namely to process fewer tokens and to improve recognition accuracy. This is in contrast to prior works which either drop tokens at the cost of accuracy, or increase accuracy whilst also increasing the computation required. First, we propose an object-guided token sampling strategy that enables us to retain a small fraction of the input tokens with minimal impact on accuracy. And second, we propose an object-aware attention module that enriches our feature representation with object information and improves overall accuracy. Our resulting framework achieves better performance when using fewer tokens than strong baselines. In particular, we match our baseline with 30%, 40%, and 60% of the input tokens on SomethingElse, Something-something v2, and Epic-Kitchens, respectively. When we use our model to process the same number of tokens as our baseline, we improve by 0.6 to 4.2 points on these datasets.

cs.CV↗
arXiv2023

Statistical Analysis of Magnetic Domain Wall Dynamics to Quantify Dzyaloshinskii-Moriya Interaction

We utilize statistical tools to analyze the magnetic domain wall dynamics in a nanostrip, which can quantify the magnitude and reveal the effects of interfacial Dzyaloshinskii-Mariya interaction. We find that there exist two peaks in the velocity frequency spectrum, the magnitude ratio of which can be used to determine the DMI strength. Our approach is validated using a collective-coordinate model, and is demonstrated to be robust against thermal noise and material impurities. Moreover, third-order cumulant and third-order time-dependent correlation function of velocity are calculated and yield valuable information regarding the asymmetry induced by DMI. Our findings offer novel and efficient analysis tools to understand physical process of domain wall dynamics under DMI and exotic magnetic phenomena.

cond-mat.mes-hall↗
arXiv2023

Do Vision-Language Pretrained Models Learn Composable Primitive Concepts?

Vision-language (VL) pretrained models have achieved impressive performance on multimodal reasoning and zero-shot recognition tasks. Many of these VL models are pretrained on unlabeled image and caption pairs from the internet. In this paper, we study whether representations of primitive concepts--such as colors, shapes, or the attributes of object parts--emerge automatically within these pretrained VL models. We propose a two-step framework, Compositional Concept Mapping (CompMap), to investigate this. CompMap first asks a VL model to generate primitive concept activations with text prompts, and then learns to construct a composition model that maps the primitive concept activations (e.g. the likelihood of black tail or red wing) to composite concepts (e.g. a red-winged blackbird). We show that a composition model can be reliably learn from ground truth primitive concepts. We thus hypothesize that if primitive concepts indeed emerge in a VL pretrained model, its primitive concept activations can be used to learn a composition model similar to the one designed by experts. We propose a quantitative metric to measure the degree of similarity, and refer to the metric as the interpretability metric. We also measure the classification accuracy when using the primitive concept activations and the learned composition model to predict the composite concepts, and refer to it as the usefulness metric. Our study reveals that state-of-the-art VL pretrained models learn primitive concepts that are highly useful for fine-grained visual recognition on the CUB dataset, and compositional generalization tasks on the MIT-States dataset. However, we observe that the learned composition models have low interpretability in our qualitative analyses. Our results reveal the limitations of existing VL models, and the necessity of pretraining objectives that encourage the acquisition of primitive concepts.

cs.CV↗
arXiv2023

Segment Any Anomaly without Training via Hybrid Prompt Regularization

We present a novel framework, i.e., Segment Any Anomaly + (SAA+), for zero-shot anomaly segmentation with hybrid prompt regularization to improve the adaptability of modern foundation models. Existing anomaly segmentation models typically rely on domain-specific fine-tuning, limiting their generalization across countless anomaly patterns. In this work, inspired by the great zero-shot generalization ability of foundation models like Segment Anything, we first explore their assembly to leverage diverse multi-modal prior knowledge for anomaly localization. For non-parameter foundation model adaptation to anomaly segmentation, we further introduce hybrid prompts derived from domain expert knowledge and target image context as regularization. Our proposed SAA+ model achieves state-of-the-art performance on several anomaly segmentation benchmarks, including VisA, MVTec-AD, MTD, and KSDD2, in the zero-shot setting. We will release the code at \href{https://github.com/caoyunkang/Segment-Any-Anomaly}{https://github.com/caoyunkang/Segment-Any-Anomaly}.

cs.CV↗
arXiv2023

End-to-End Spatio-Temporal Action Localisation with Video Transformers

The most performant spatio-temporal action localisation models use external person proposals and complex external memory banks. We propose a fully end-to-end, purely-transformer based model that directly ingests an input video, and outputs tubelets -- a sequence of bounding boxes and the action classes at each frame. Our flexible model can be trained with either sparse bounding-box supervision on individual frames, or full tubelet annotations. And in both cases, it predicts coherent tubelets as the output. Moreover, our end-to-end model requires no additional pre-processing in the form of proposals, or post-processing in terms of non-maximal suppression. We perform extensive ablation experiments, and significantly advance the state-of-the-art results on four different spatio-temporal action localisation benchmarks with both sparse keyframes and full tubelet annotations.

cs.CV↗
arXiv2023

REVEAL: Retrieval-Augmented Visual-Language Pre-Training with Multi-Source Multimodal Knowledge Memory

In this paper, we propose an end-to-end Retrieval-Augmented Visual Language Model (REVEAL) that learns to encode world knowledge into a large-scale memory, and to retrieve from it to answer knowledge-intensive queries. REVEAL consists of four key components: the memory, the encoder, the retriever and the generator. The large-scale memory encodes various sources of multimodal world knowledge (e.g. image-text pairs, question answering pairs, knowledge graph triplets, etc) via a unified encoder. The retriever finds the most relevant knowledge entries in the memory, and the generator fuses the retrieved knowledge with the input query to produce the output. A key novelty in our approach is that the memory, encoder, retriever and generator are all pre-trained end-to-end on a massive amount of data. Furthermore, our approach can use a diverse set of multimodal knowledge sources, which is shown to result in significant gains. We show that REVEAL achieves state-of-the-art results on visual question answering and image captioning.

cs.CV↗
arXiv2023

Spin Diffusion in Spin Glasses Require Two Magnetic Variables, $\vec{M}$ and $\vec{m}$

Experiment has established that spin-glasses can support a steady-state spin current $\vec{j}_{i}$. However, the accepted theory of spin glass dynamics permits oscillations but no steady-state spin current. Onsager's irreversible thermodynamics implies that the spin current is proportional to the gradient of a magnetization. We argue, however, that the magnon distribution function associated with the local equilibrium magnetization $\vec{M}$ cannot diffuse because it represents $10^{23}$ variables. We therefore invoke the non-equilibrium magnetization $\vec{m}$, which in spintronics is called the {\it spin accumulation}. Applying the theory of irreversible thermodynamics we indeed find that it predicts spin diffusion, and we consider other experimental consequences of the theory, including a wavelength-dependent coupling between the reactive and the diffusive degrees of freedom.

cond-mat.mes-hall↗
arXiv2023

Reconfigurable Intelligent Surface-aided Secret Key Generation in Multi-Cell Systems

Physical-layer key generation (PKG) exploits the reciprocity and randomness of wireless channels to generate a symmetric key between two legitimate communication ends. However, in multi-cell systems, PKG suffers from severe pilot contamination due to the reuse of pilots in different cells. In this paper, we invoke multiple reconfigurable intelligent surfaces (RISs) for adaptively shaping the environment and enhancing the PKG performance. To this end, we formulate an optimization problem to maximize the weighted sum key rate (WSKR) by jointly optimizing the precoding matrices at the base stations (BSs) and the phase shifts at the RISs. For addressing the non-convexity of the problem, we derive an upper bound of the WSKR and prove its tightness. To tackle the upper bound maximization problem, we apply an alternating optimization (AO)-based algorithm to divide the joint optimization into two sub-problems. We apply the Lagrangian dual approach based on the Karush-Kuhn-Tucker (KKT) conditions for the sub-problem of precoding matrices and adopt a projected gradient ascent (PGA) algorithm for the sub-problem of phase shifts. Simulation results confirm the near-optimal performance of the proposed algorithm and the effectiveness of RISs for improving the WSKR via mitigating pilot contamination.

cs.IT↗
arXiv2023

Blockchain Based Spectrum Sharing

Spectrum sharing has long been considered as method to improve spectrum resource utilization. Centralized geolocation database approach has been accepted globally for commercial applications. Recently blockchain has been considered as a platform to support spectrum sharing in a distributed manner. Like other commodities, spectrum or right of spectrum usage can be exchanged through blockchain. However, this leads to changes of the location where a particular frequency channel is utilized which brings potential interference. In this paper, we present the interference based consensus and smart contract for blockchain based spectrum sharing. We will also introduce the other aspects of blockchain such as cross train transaction, cross tier spectrum coordination and incentive mechanism that need to be studied for spectrum sharing applications.

cs.NI↗
arXiv2023

DEJA VU: Continual Model Generalization For Unseen Domains

In real-world applications, deep learning models often run in non-stationary environments where the target data distribution continually shifts over time. There have been numerous domain adaptation (DA) methods in both online and offline modes to improve cross-domain adaptation ability. However, these DA methods typically only provide good performance after a long period of adaptation, and perform poorly on new domains before and during adaptation - in what we call the "Unfamiliar Period", especially when domain shifts happen suddenly and significantly. On the other hand, domain generalization (DG) methods have been proposed to improve the model generalization ability on unadapted domains. However, existing DG works are ineffective for continually changing domains due to severe catastrophic forgetting of learned knowledge. To overcome these limitations of DA and DG in handling the Unfamiliar Period during continual domain shift, we propose RaTP, a framework that focuses on improving models' target domain generalization (TDG) capability, while also achieving effective target domain adaptation (TDA) capability right after training on certain domains and forgetting alleviation (FA) capability on past domains. RaTP includes a training-free data augmentation module to prepare data for TDG, a novel pseudo-labeling mechanism to provide reliable supervision for TDA, and a prototype contrastive alignment algorithm to align different domains for achieving TDG, TDA and FA. Extensive experiments on Digits, PACS, and DomainNet demonstrate that RaTP significantly outperforms state-of-the-art works from Continual DA, Source-Free DA, Test-Time/Online DA, Single DG, Multiple DG and Unified DA&DG in TDG, and achieves comparable TDA and FA capabilities.

cs.LG↗
arXiv2023

Comparing Trajectory and Vision Modalities for Verb Representation

Three-dimensional trajectories, or the 3D position and rotation of objects over time, have been shown to encode key aspects of verb semantics (e.g., the meanings of roll vs. slide). However, most multimodal models in NLP use 2D images as representations of the world. Given the importance of 3D space in formal models of verb semantics, we expect that these 2D images would result in impoverished representations that fail to capture nuanced differences in meaning. This paper tests this hypothesis directly in controlled experiments. We train self-supervised image and trajectory encoders, and then evaluate them on the extent to which each learns to differentiate verb concepts. Contrary to our initial expectations, we find that 2D visual modalities perform similarly well to 3D trajectories. While further work should be conducted on this question, our initial findings challenge the conventional wisdom that richer environment representations necessarily translate into better representation learning for language.

cs.CV↗
arXiv2023

Steerable Equivariant Representation Learning

Pre-trained deep image representations are useful for post-training tasks such as classification through transfer learning, image retrieval, and object detection. Data augmentations are a crucial aspect of pre-training robust representations in both supervised and self-supervised settings. Data augmentations explicitly or implicitly promote invariance in the embedding space to the input image transformations. This invariance reduces generalization to those downstream tasks which rely on sensitivity to these particular data augmentations. In this paper, we propose a method of learning representations that are instead equivariant to data augmentations. We achieve this equivariance through the use of steerable representations. Our representations can be manipulated directly in embedding space via learned linear maps. We demonstrate that our resulting steerable and equivariant representations lead to better performance on transfer learning and robustness: e.g. we improve linear probe top-1 accuracy by between 1% to 3% for transfer; and ImageNet-C accuracy by upto 3.4%. We further show that the steerability of our representations provides significant speedup (nearly 50x) for test-time augmentations; by applying a large number of augmentations for out-of-distribution detection, we significantly improve OOD AUC on the ImageNet-C dataset over an invariant representation.

cs.CV↗
arXiv2023

Eliminating Leakage in Volatile Memory with Anti-Ferroelectric Transistors

Cache serves as a temporary data memory module in many general-purpose processors and domain-specific accelerators. Its density, power, speed, and reliability play a critical role in enhancing the overall system performance and quality of service. Conventional volatile memories, including static random-access memory (SRAM) and embedded dynamic random-access memory (eDRAM) in the complementary metal-oxide-semiconductor technology, have high performance and good reliability. However, the inherent leakage in both SRAM and eDRAM hinders further improvement towards smaller feature sizes and higher energy efficiency. Although the emerging nonvolatile memories can eliminate the leakage efficiently, the penalties of lower speed and degraded reliability are significant. This article reveals a new opportunity towards leakage-free volatile static memory beyond the known paradigms of existing volatile and nonvolatile memories. By engineering a double-well energy landscape with the assistance of a clamping voltage bias, leakage-free and refresh-free state retention of volatile memory is achieved for the first time. This new memory is highlighted by both the ultra-low leakage of nonvolatile memories and the speed, energy, and reliability advantages of volatile memories. A proof-of-concept memory is demonstrated using in-house anti-ferroelectric field-effect transistors (AFeFETs), delivering an extrapolated endurance of about 1012 cycles, a retention time of over 10 years, and no subthreshold channel leakage current. Such a new concept of AFeFET-based memory enables an improved balance between density, power, and reliability beyond all existing memory solutions.

cs.ET↗
arXiv2022

Computational Associative Memory with Amorphous InGaZnO Channel 3D NAND-Compatible FG Transistors

3D NAND enables continuous NAND density and cost scaling beyond conventional 2D NAND. However, its poly-Si channel suffers from low mobility, large device variations, and instability caused by grain boundaries. Here, we overcome these drawbacks by introducing an amorphous indium-gallium-zinc-oxide (a-IGZO) channel, which has the advantages of ultra-low OFF current, back-end-of-line compatibility, higher mobility and better uniformity than poly-Si, and free of grain boundaries due to the amorphous nature. Ultra-scaled floating-gate (FG) transistors with a channel length of 60 nm are reported, achieving the highest ON current of 127 uA/um among all reported a-IGZO-based flash devices for high-density, low-power, and high-performance 3D NAND applications. Furthermore, a non-volatile and area-efficient ternary content-addressable memory (TCAM) with only two a-IGZO FG transistors is experimentally demonstrated. Array-level simulations using experimentally calibrated models show that this design achieves at least 240x array-size scalability and 2.7-fold reduction in search energy than 16T-CMOS, 2T2R, and 2FeFET TCAMs.

cond-mat.mes-hall↗
arXiv2022

Many-body slow quench dynamics and nonadiabatic characterization of topological phases

Previous studies have shown that the bulk topology of single-particle systems can be captured by the band inversion surface or by the spin inversion surface emerged on the time-averaged spin polarization. Most of the studies, however, are based on the single-particle picture even though the systems are fermionic and of multi-bands. Here, we study the many-body quench dynamics of topological systems with all the valence bands fully occupied, and show that the concepts of band inversion surface and spin inversion surface are still valid. More importantly, the many-body quench dynamics is shown to be reduced to a nontrivial three-level Landau-Zener model, which can be solved exactly. Based on the analytical results, the topological spin texture revealed by the time-averaged spin polarization can be applied to characterize the bulk topology and thus provides a direct comparison for future experiments.

cond-mat.str-el↗