Multimodal Visual Object Tracking (VOT) has recently gained significant attention due to its robustness. Early research focused on fully fine-tuning RGB-based trackers, which was inefficient and lacked generalized representation due to the scarcity of multimodal data. Therefore, recent studies have utilized prompt tuning to transfer pre-trained RGB-based trackers to multimodal data. However, the modality gap limits pre-trained knowledge recall, and the dominance of the RGB modality persists, preventing the full utilization of information from other modalities. To address these issues, we propose a novel symmetric multimodal tracking framework called SDSTrack. We introduce lightweight adaptation for efficient fine-tuning, which directly transfers the feature extraction ability from RGB to other domains with a small number of trainable parameters and integrates multimodal features in a balanced, symmetric manner. Furthermore, we design a complementary masked patch distillation strategy to enhance the robustness of trackers in complex environments, such as extreme weather, poor imaging, and sensor failure. Extensive experiments demonstrate that SDSTrack outperforms state-of-the-art methods in various multimodal tracking scenarios, including RGB+Depth, RGB+Thermal, and RGB+Event tracking, and exhibits impressive results in extreme conditions. Our source code is available at https://github.com/hoqolo/SDSTrack.
Integrated sensing and communication (ISAC) has recently attracted tremendous attention from both academia and industry, being envisioned as a key part of the standards for the sixth-generation (6G) cellular network. A key challenge of 6G-oriented ISAC lies in how to perform ubiquitous sensing based on the communication signals and devices. Previous works have made great progresses on studying the signal waveform design that leads to optimal communication-sensing performance tradeoff. In this article, we aim to focus on issues arising from the exploitation of the communication devices for sensing in 6G network. Particularly, we will discuss about how to leverage various nodes available in the cellular network as anchors to perform ubiquitous sensing. On one hand, the base stations (BSs) will be the most important anchors in the future 6G ISAC network, since they can generate/process radio signals with high range/angle resolutions, and their positions are precisely known. Correspondingly, we will first study the BS-based sensing technique. On the other hand, the BSs alone may not enable ubiquitous sensing, since they cannot cover all the places with strong line-of-sight (LOS) links. This motivates us to investigate the possibility of using other nodes that are with higher density in the network to act as the anchors. Along this line, we are interested in two types of new anchors - user equipments (UEs) and reconfigurable intelligent surfaces (RISs). This paper will shed light on the opportunities and challenges brought by UE-assisted sensing and RIS-assisted sensing. Our goal is to devise a novel 6G-oriented sensing architecture where BSs, UEs, and RISs can work together to provide ubiquitous sensing services.
In the field of multi-class anomaly detection, reconstruction-based methods derived from single-class anomaly detection face the well-known challenge of ``learning shortcuts'', wherein the model fails to learn the patterns of normal samples as it should, opting instead for shortcuts such as identity mapping or artificial noise elimination. Consequently, the model becomes unable to reconstruct genuine anomalies as normal instances, resulting in a failure of anomaly detection. To counter this issue, we present a novel unified feature reconstruction-based anomaly detection framework termed RLR (Reconstruct features from a Learnable Reference representation). Unlike previous methods, RLR utilizes learnable reference representations to compel the model to learn normal feature patterns explicitly, thereby prevents the model from succumbing to the ``learning shortcuts'' issue. Additionally, RLR incorporates locality constraints into the learnable reference to facilitate more effective normal pattern capture and utilizes a masked learnable key attention mechanism to enhance robustness. Evaluation of RLR on the 15-category MVTec-AD dataset and the 12-category VisA dataset shows superior performance compared to state-of-the-art methods under the unified setting. The code of RLR will be publicly available.
Few-shot anomaly detection (FSAD) is essential in industrial manufacturing. However, existing FSAD methods struggle to effectively leverage a limited number of normal samples, and they may fail to detect and locate inconspicuous anomalies in the spatial domain. We further discover that these subtle anomalies would be more noticeable in the frequency domain. In this paper, we propose a Dual-Path Frequency Discriminators (DFD) network from a frequency perspective to tackle these issues. Specifically, we generate anomalies at both image-level and feature-level. Differential frequency components are extracted by the multi-frequency information construction module and supplied into the fine-grained feature construction module to provide adapted features. We consider anomaly detection as a discriminative classification problem, wherefore the dual-path feature discrimination module is employed to detect and locate the image-level and feature-level anomalies in the feature space. The discriminators aim to learn a joint representation of anomalous features and normal features in the latent space. Extensive experiments conducted on MVTec AD and VisA benchmarks demonstrate that our DFD surpasses current state-of-the-art methods. Source code will be available.
Recent advancements in foundation models have yielded impressive performance across a wide range of tasks. Meanwhile, for specific applications, practitioners have been developing specialized application models. To enjoy the benefits of both kinds of models, one natural path is to transfer the knowledge in foundation models into specialized application models, which are generally more efficient for serving. Techniques from knowledge distillation may be applied here, where the application model learns to mimic the foundation model. However, specialized application models and foundation models have substantial gaps in capacity, employing distinct architectures, using different input features from different modalities, and being optimized on different distributions. These differences in model characteristics lead to significant challenges for distillation methods. In this work, we propose creating a teaching committee comprising both foundation model teachers and complementary teachers. Complementary teachers possess model characteristics akin to the student's, aiming to bridge the gap between the foundation model and specialized application models for a smoother knowledge transfer. Further, to accommodate the dissimilarity among the teachers in the committee, we introduce DiverseDistill, which allows the student to understand the expertise of each teacher and extract task knowledge. Our evaluations demonstrate that adding complementary teachers enhances student performance. Finally, DiverseDistill consistently outperforms baseline distillation methods, regardless of the teacher choices, resulting in significantly improved student performance.
Recent advancements in foundation models have yielded impressive performance across a wide range of tasks. Meanwhile, for specific applications, practitioners have been developing specialized application models. To enjoy the benefits of both kinds of models, one natural path is to transfer the knowledge in foundation models into specialized application models, which are generally more efficient for serving. Techniques from knowledge distillation may be applied here, where the application model learns to mimic the foundation model. However, specialized application models and foundation models have substantial gaps in capacity, employing distinct architectures, using different input features from different modalities, and being optimized on different distributions. These differences in model characteristics lead to significant challenges for distillation methods. In this work, we propose creating a teaching committee comprising both foundation model teachers and complementary teachers. Complementary teachers possess model characteristics akin to the student's, aiming to bridge the gap between the foundation model and specialized application models for a smoother knowledge transfer. Further, to accommodate the dissimilarity among the teachers in the committee, we introduce DiverseDistill, which allows the student to understand the expertise of each teacher and extract task knowledge. Our evaluations demonstrate that adding complementary teachers enhances student performance. Finally, DiverseDistill consistently outperforms baseline distillation methods, regardless of the teacher choices, resulting in significantly improved student performance.
Fine-tuning is becoming widely used for leveraging the power of pre-trained foundation models in new downstream tasks. While there are many successes of fine-tuning on various tasks, recent studies have observed challenges in the generalization of fine-tuned models to unseen distributions (i.e., out-of-distribution; OOD). To improve OOD generalization, some previous studies identify the limitations of fine-tuning data and regulate fine-tuning to preserve the general representation learned from pre-training data. However, potential limitations in the pre-training data and models are often ignored. In this paper, we contend that overly relying on the pre-trained representation may hinder fine-tuning from learning essential representations for downstream tasks and thus hurt its OOD generalization. It can be especially catastrophic when new tasks are from different (sub)domains compared to pre-training data. To address the issues in both pre-training and fine-tuning data, we propose a novel generalizable fine-tuning method LEVI, where the pre-trained model is adaptively ensembled layer-wise with a small task-specific model, while preserving training and inference efficiencies. By combining two complementing models, LEVI effectively suppresses problematic features in both the fine-tuning data and pre-trained model and preserves useful features for new tasks. Broad experiments with large language and vision models show that LEVI greatly improves fine-tuning generalization via emphasizing different views from fine-tuning data and pre-trained features.
Multi-target detection is one of the primary tasks in radar-based localization and sensing, typically built on phased array antennas. However, the bulky hardware in the phased array restricts its potential for enhancing detection accuracy, since the cost and power of the phased array can become unaffordable as its physical aperture scales up to pursue higher beam shaping capabilities. To resolve this issue, we propose a radar system enabled by reconfigurable holographic surfaces (RHSs), a novel meta-surface antenna composed of meta-material elements with cost-effective and power-efficient hardware, which performs multi-target detection in an adaptive manner. Different from the phase-control structure in the phased array, the RHS is able to apply beamforming by controlling the radiation amplitudes of its elements. Consequently, traditional beamforming schemes designed for phased arrays cannot be directly applied to RHSs due to this structural difference. To tackle this challenge, a waveform and amplitude optimization algorithm (WAOA) is designed to jointly optimize the radar waveform and RHS amplitudes in order to improve the detection accuracy. Simulation results reveal that the proposed RHS-enabled radar increases the probability of detection by 0.13 compared to phased array radars when six iterations of adaptive detection are performed given the same hardware cost.
Industrial anomaly detection is generally addressed as an unsupervised task that aims at locating defects with only normal training samples. Recently, numerous 2D anomaly detection methods have been proposed and have achieved promising results, however, using only the 2D RGB data as input is not sufficient to identify imperceptible geometric surface anomalies. Hence, in this work, we focus on multi-modal anomaly detection. Specifically, we investigate early multi-modal approaches that attempted to utilize models pre-trained on large-scale visual datasets, i.e., ImageNet, to construct feature databases. And we empirically find that directly using these pre-trained models is not optimal, it can either fail to detect subtle defects or mistake abnormal features as normal ones. This may be attributed to the domain gap between target industrial data and source data.Towards this problem, we propose a Local-to-global Self-supervised Feature Adaptation (LSFA) method to finetune the adaptors and learn task-oriented representation toward anomaly detection.Both intra-modal adaptation and cross-modal alignment are optimized from a local-to-global perspective in LSFA to ensure the representation quality and consistency in the inference stage.Extensive experiments demonstrate that our method not only brings a significant performance boost to feature embedding based approaches, but also outperforms previous State-of-The-Art (SoTA) methods prominently on both MVTec-3D AD and Eyecandies datasets, e.g., LSFA achieves 97.1% I-AUROC on MVTec-3D, surpass previous SoTA by +3.4%.
To realize holographic communications, a potential technology for spectrum efficiency improvement in the future sixth-generation (6G) network, antenna arrays inlaid with numerous antenna elements will be deployed. However, the increase in antenna aperture size makes some users lie in the Fresnel region, leading to the hybrid near-field and far-field communication mode, where the conventional far-field channel estimation methods no longer work well. To tackle the above challenge, this paper considers channel estimation in a hybrid-field multipath environment, where each user and each scatterer can be in either the far-field or the near-field region. First, a joint angular-polar domain channel transform is designed to capture the hybrid-field channel's near-field and far-field features. We then analyze the power diffusion effect in the hybrid-field channel, which indicates that the power corresponding to one near-field (far-field) path component of the multipath channel may spread to far-field (near-field) paths and causes estimation error. We design a novel power-diffusion-based orthogonal matching pursuit channel estimation algorithm (PD-OMP). It can eliminate the prior knowledge requirement of path numbers in the far field and near field, which is a must in other OMP-based channel estimation algorithms. Simulation results show that PD-OMP outperforms current hybrid-field channel estimation methods.