Pseudo-labeling is a cornerstone of Unsupervised Domain Adaptation (UDA), yet the scarcity of High-Confidence Pseudo-Labeled Target Domain Samples (hcpl-tds) often leads to inaccurate cross-domain statistical alignment, causing DA failures. To address this challenge, we propose Noise Optimized Conditional Diffusion for Domain Adaptation (NOCDDA), which seamlessly integrates the generative capabilities of conditional diffusion models with the decision-making requirements of DA to achieve task-coupled optimization for efficient adaptation. For robust cross-domain consistency, we modify the DA classifier to align with the conditional diffusion classifier within a unified optimization framework, enabling forward training on noise-varying cross-domain samples. Furthermore, we argue that the conventional N(0,I) initialization in diffusion models often generates class-confused hcpl-tds, compromising discriminative DA. To resolve this, we introduce a class-aware noise optimization strategy that refines sampling regions for reverse class-specific hcpl-tds generation, effectively enhancing cross-domain alignment. Extensive experiments across 5 benchmark datasets and 29 DA tasks demonstrate significant performance gains of NOCDDA over 31 state-of-the-art methods, validating its robustness and effectiveness.

Road detection using camera-LiDAR fusion is a critical yet challenging task, as it combines the semantic richness of camera data with the depth precision of LiDAR, enabling robust performance under diverse conditions. However, this task faces two key challenges: (1) the limited adoption of LiDAR sensors due to their high cost, and (2) the scarcity of labeled data in new, unseen testing datasets, which hampers the generalization of supervised algorithms. To address the first challenge, we developed a Pseudo-LiDAR (PL) generation mechanism that synthesizes depth information to augment existing datasets, reducing reliance on real LiDAR sensors. For the second challenge, we proposed an unsupervised method combining Superpixel-assisted segmentation with adversarial Domain Adaptation (SDA). By aligning cross-domain features, this method enhances the network’s ability to accurately distinguish road boundaries on unseen data, enabling effective unsupervised road detection. Our complete framework, PL-SDA-Net, was trained on the KITTI Road dataset, which includes both labeled RGB and LiDAR data, and validated on two datasets without LiDAR data, Cityscapes and CamVid. Experimental results demonstrate that our method achieves competitive performance across various metrics, outperforming several state-of-the-art techniques.

In domain adaptation (DA), the effectiveness of deep learning-based models is often constrained by batch learning strategies that fail to fully apprehend the global statistical and geometric characteristics of data distributions. Addressing this gap, we introduce "Global Awareness Enhanced Domain Adaptation" (GAN-DA), a novel approach that transcends traditional batch-based limitations. GAN-DA integrates a unique predefined feature representation (PFR) to facilitate the alignment of cross-domain distributions, thereby achieving a comprehensive global statistical awareness. This representation is innovatively expanded to encompass orthogonal and common feature aspects, which enhances the unification of global manifold structures and refines decision boundaries for more effective DA. Our extensive experiments, encompassing 27 diverse cross-domain image classification tasks, demonstrate GAN-DA's remarkable superiority, outperforming 24 established DA methods by a significant margin. Furthermore, our in-depth analyses shed light on the decision-making processes, revealing insights into the adaptability and efficiency of GAN-DA. This approach not only addresses the limitations of existing DA methodologies but also sets a new benchmark in the realm of domain adaptation, offering broad implications for future research and applications in this field.

The ground communication system for Unmanned Aerial Vehicles (UAVs) plays a crucial role in various field, including warfare, agriculture, energy, fire monitoring and transportation . As the number of UAVs integrated into ground communication station continues to grow, there is a corresponding increase in communication requirements. The conventional network architecture of UAV ground communication stations is no longer sufficient to meet the diverse demands of various combat scenarios. To address this challenge, this paper explores existing design issues related to the network architecture of UAV ground command systems. It puts forth a novel network architecture tailored for extensive UAV-to-ground communication, grounded in the concept of network slicing. Additionally, the paper outlines key technologies and supplementary mechanisms required to implement this envisioned network architecture.

In leveraging manifold learning in domain adaptation (DA), graph embedding-based DA methods have shown their effectiveness in preserving data manifold through the Laplace graph. However, current graph embedding DA methods suffer from two issues: 1). they are only concerned with preservation of the underlying data structures in the embedding and ignore sub-domain adaptation, which requires taking into account intra-class similarity and inter-class dissimilarity, thereby leading to negative transfer; 2). manifold learning is proposed across different feature/label spaces separately, thereby hindering unified comprehensive manifold learning. In this paper, starting from our previous DGA-DA, we propose a novel DA method, namely A ttention R egularized Laplace G raph-based D omain A daptation (ARG-DA), to remedy the aforementioned issues. Specifically, by weighting the importance across different sub-domain adaptation tasks, we propose the A ttention R egularized Laplace Graph for class aware DA, thereby generating the attention regularized DA. Furthermore, using a specifically designed FEEL strategy, our approach dynamically unifies alignment of the manifold structures across different feature/label spaces, thus leading to comprehensive manifold learning. Comprehensive experiments are carried out to verify the effectiveness of the proposed DA method, which consistently outperforms the state of the art DA methods on 7 standard DA benchmarks, i.e., 37 cross-domain image classification tasks including object, face, and digit images. An in-depth analysis of the proposed DA method is also discussed, including sensitivity, convergence, and robustness.

Person re-identification (re-ID) which aims to recognize a pedestrian observed by non-overlapping cameras is a challenging task due to high variance between images from different viewpoints. Although remarkable progresses on research of re-ID had been obtained via leveraging the merits of deep learning framework through sufficient quantity training on a large amount of well labeled data, whereas, in real scenarios, re-ID generally suffers from lacking of well labeled training data. In this paper, we propose an attention-based model with attribute classification (AMAC) to facilitate a well trained model transferring across different data domains, which further enables an efficient cross-domain video-based person re-ID. Specifically, an attention-based sub-network is proposed for deep insight into the quality variations of local parts, hence, different local parts are cooperated with different weights to avoid the heavy occlusions or the cluttered background in datasets. Moreover, we introduce a transferred attribute classification sub-network to extract attribute-semantic features of any new target datasets without the requirement for new training attribute labels which are costly to annotate. Attribute-semantic features can be considered as valuable complementary information for person reidentification since they are robust to illumination varieties and different viewpoints across cameras. Due to the large gap between different datasets, we finetune each sub-network with pseudo labels on the target datasets respectively to strengthen the original model trained on other labeled datasets. Extensive comparable evaluations demonstrate the superiority of our AMAC in solving cross-domain person re- ID task on two benchmarks including PRID-2011 and iLIDS-VID.

Domain adaptation (DA) aims to generalize a learning model across training and testing data despite the mismatch of their data distributions. In light of a theoretical estimation of the upper error bound, we argue, in this article, that an effective DA method for classification should: 1) search a shared feature subspace where the source and target data are not only aligned in terms of distributions as most state-of-the-art DA methods do but also discriminative in that instances of different classes are well separated and 2) account for the geometric structure of the underlying data manifold when inferring data labels on the target domain. In comparison with a baseline DA method which only cares about data distribution alignment between source and target, we derive three different DA models for classification, namely, close yet discriminative DA (CDDA), geometry-aware DA (GA-DA), and discriminative and GA-DA (DGA-DA), to highlight the contribution of CDDA based on 1), GA-DA based on 2), and, finally, DGA-DA implementing jointly 1) and 2). Using both the synthetic and real data, we show the effectiveness of the proposed approach which consistently outperforms the state-of-the-art DA methods over 49 image classification DA tasks through eight popular benchmarks. We further carry out an in-depth analysis of the proposed DA method in quantifying the contribution of each term of our DA model and provide insights into the proposed DA methods in visualizing both real and synthetic data.

Visual tracking has recently gained a great advance with the use of the convolutional neural network (CNN). Usually, existing CNN‐based trackers exploit the features from a single layer or a certain combination of multiple layers. However, these features only characterise an object from an invariable aspect and cannot adapt to scene variation, which limits the performance of such trackers. To overcome this limitation, the authors study the problem from a new perspective and propose a novel convolutional layer selection method. To obtain robust appearance representation, they investigate the advantages of features extracted from different convolutional layers. To determine the correctness of the tracking prediction and updated model, they design a verification mechanism based on historical retrospect, which can estimate the deviation for each layer by bidirectionally locating the target. Meanwhile, the deviation works as the layer‐wise selection criteria. Extensive evaluations on the OTB‐2013, visual object tracking (VOT)‐2016 and VOT‐2017 benchmarks demonstrate that the proposed tracker performs favourably against several state‐of‐the‐art trackers.

Local binary pattern (LBP) is one of the most successful feature descriptors. However, LBP and its variants have not been as successful as other feature descriptors in video anomaly detection (VAD). This is because LBP and its variants are mainly designed for spatial texture analysis. Although the volume LBP (VLBP) and the LBP-three orthogonal planes (LPB-TOP) have the capability of describing dynamic texture, they are seldom used as descriptors for VAD because 1) both VLBP and LBP-TOP are more suitable for natural scenes with rich dynamic textures, but sensitive to noise in the scenes with less dynamic textures, 2) the combination of motion and appearance not only limits their capability of motion characterizing but also brings the irrelevant appearance information such as background, and 3) high dimensionality is another drawback. In this paper, a new variant of the LBP called the squirrel-cage LBP (SCLBP) is proposed for VAD. By imitating the structure of squirrel cage rotor, the proposed SCLBP can be regarded as a stretched LBP in temporal direction and has two distinct features: 1) it is computed at vector-wise, rather than at pixel-wise (i.e., for the central vector with respect to its surrounding parallel vectors), and 2) the sign between two vectors is determined by the angle-based thresholding scheme. The SCLBP can effectively encode the motion information and is insensitive to noise and irrelevant disturbances caused by dynamic background and illumination change. To the best of our knowledge, The SCLBP is the first variant of the LBP specially designed for motion characterizing. The SCLBP has great flexibility, extendibility, and low dimensionality (only one-third of the LBP-TOP descriptor). The effectiveness of the proposed SCLBP descriptor is demonstrated on different public data sets and compared with the other dominant descriptors and state-of-the-art approaches in VAD.

This letter presents a novel visual tracking algorithm by using Incremental Non-negative Matrix Factorization (INMF) and dual ℓ1-norm constraints. Firstly, we introduce one ℓ1regularization into the NMF reconstruction, which enables appearance model to tolerate different noises to some extent. Meanwhile, we enforce another ℓ1regularization on the projection coefficients when using iterative operators to obtain NMF basis vectors for the effective tracking. Secondly, to obtain the sparse error and projection coefficient matrice, we present an iterative algorithm to solve the optimal problem, which ensures the representation is more robust. Finally, we take partial occlusion into construct likelihood function, and combined with INMF learning to update appearance model for alleviating tracking drift. Experimental results compared with the state-of-the-art tracking methods demonstrate the proposed algorithm achieves favorable performance when the object undergoes large occlusion, motion blur and illumination changes.