Referring multi-object tracking (RMOT) aims to track every instance in a video that matches a given language expression. Despite the recent integration of multimodal large language models (MLLMs) to enhance generalization, existing methods predominantly relegate them to the role of caption generators, necessitating external modules for final decision-making. This paradigm not only introduces extra latency but also severely underutilizes the inherent vision-language alignment capabilities of MLLMs. To address these limitations, we propose YesTrack, a novel two-stage RMOT method that reformulates referring as a discriminative task, directly leveraging MLLMs for Yes/No verification without explicit text generation. To further enhance the reliability and efficiency of this MLLM-based verification, we introduce two lightweight temporal consistency constraints: Temporal Confidence Prior (TCP) and Temporal Reference Propagation (TRP). We further validate the generality of this discriminative paradigm by proposing YesTrack-MOT, a straightforward yet highly effective instantiation for generic multi-object tracking (MOT). Experiments on Refer-KITTI and Refer-KITTI-V2 show that YesTrack significantly outperforms existing state-of-the-art methods while maintaining high efficiency, even when implemented with the smallest variant of Qwen3-VL. Code is released at https://github.com/ggbondrighthere24/YesTrack.

Diffusion-based Real-World Image Super-Resolution (Real-ISR) achieves impressive perceptual quality but suffers from high computational costs due to iterative sampling. While recent distillation approaches leveraging large-scale Text-to-Image (T2I) priors have enabled one-step generation, they are typically hindered by prohibitive parameter counts and the inherent capability bounds imposed by teacher models. As a lightweight alternative, Consistency Models offer efficient inference but struggle with two critical limitations: the accumulation of consistency drift inherent to transitive training, and a phenomenon we term "Geometric Decoupling"— where the generative trajectory achieves pixel-wise alignment yet fails to preserve structural coherence. To address these challenges, we propose GTASR (Geometric Trajectory Alignment Super-Resolution), a {simple yet effective} consistency training paradigm for Real-ISR. Specifically, we introduce a Trajectory Alignment (TA) strategy to rectify the tangent vector field via full-path projection, and a Dual-Reference Structural Rectification (DRSR) mechanism to enforce strict structural constraints. Extensive experiments verify that GTASR delivers superior performance over representative baselines while maintaining minimal latency.

Diffusion-based real-world image super-resolution (Real-ISR) has achieved remarkable perceptual quality; however, directly super-resolving images to 4K remains limited by extreme memory consumption. Consequently, prior methods adopt patch-based inference, sacrificing global context and introducing semantic confusion, spatial inconsistency, and severe latency. We propose OP4KSR, a one-step patch-free 4K SR approach built upon the powerful Flux backbone. By leveraging the extreme-compression F16 VAE, OP4KSR makes 4K SR inference tractable under practical GPU budgets, preserving global spatial-semantic coherence while enabling highly efficient inference. However, adapting this one-step architecture intrinsically triggers severe periodic artifacts. We trace this to a RoPE base frequency allocation mismatch and intra-token spatial ambiguity, both exacerbated by the lack of iterative refinement. To suppress these artifacts, we couple RoPE base frequency rescaling (RFR) with an autocorrelation-based periodicity loss (ℒ_AP). Furthermore, we curate a dedicated training dataset alongside three benchmarks (one synthetic and two real-world) to advance 4K SR research. Extensive experiments demonstrate that OP4KSR achieves competitive perceptual quality with efficient inference, generating a 4096×4096 output in only 5.75 seconds on a single NVIDIA H20 GPU.

Text-to-Image Person Retrieval (TIPR) aims to retrieve pedestrian images with a given natural-language description. It remains highly challenging due to the inherent ambiguity in cross-modal alignment: existing models often struggle to capture fine-grained correspondences, and their understanding of detailed pedestrian attributes is typically confined to partial or coarse cues, leading to mismatched or erroneous retrieval results.To overcome this challenge, we propose CECA, a Conversation-Enhanced Cross-modal Alignment framework. CECA strengthens the attribute correspondence between textual and visual modalities through multimodal large language models (MLLMs)-guided dialogue, enhances detailed cross-modal matching via a Bidirectional Correlation Matching (BCM) mechanism, and stabilizes optimization with a Confidence-Aware Weighting Loss (CAWL) that reduces the impact of low-quality conversational responses. Extensive experiments on three public benchmarks demonstrate the superior performance and strong generalization ability of our approach.

Video super-resolution (VSR) is used to compose high-resolution (HR) video from low-resolution video. Recently, the deformable alignment-based VSR methods are becoming increasingly popular. In these methods, the features extracted from video are aligned to eliminate the motion error targeting high super-resolution (SR) quality. However, these methods often suffer from misalignment and the lack of enough temporal information to compose HR frames, which accordingly induce artifacts in the SR result. In this article, we design a deep VSR network (DVSRNet) based on the proposed progressive deformable alignment (PDA) module and temporal-sparse enhancement (TSE) module. Specifically, the PDA module is designed to accurately align features and to eliminate artifacts via the bidirectional information propagation. The TSE module is constructed to further eliminate artifacts and to generate clear details for the HR frame. In addition, we construct a lightweight deep optical flow network (OFNet) to obtain the bidirectional optical flows for the implementation of the PDA module. Moreover, two new loss functions are designed for our proposed method. The first one is adopted in OFNet and the second one is constructed to guarantee the generation of sharp and clear details for the HR frames. The experimental results demonstrate that our method performs better than the state-of-the-art methods.

Multi-object tracking (MOT) emerges as a pivotal and highly promising branch in the field of computer vision. Classical closed-vocabulary MOT (CV-MOT) methods aim to track objects of predefined categories. Recently, some open-vocabulary MOT (OV-MOT) methods have successfully addressed the problem of tracking unknown categories. However, we found that the CV-MOT and OV-MOT methods each struggle to excel in the tasks of the other. In this paper, we present a unified framework, Associate Everything Detected (AED), that simultaneously tackles CV-MOT and OV-MOT by integrating with any off-the-shelf detector and supports unknown categories. Different from existing tracking-by-detection MOT methods, AED gets rid of prior knowledge (e.g. motion cues) and relies solely on highly robust feature learning to handle complex trajectories in OV-MOT tasks while keeping excellent performance in CV-MOT tasks. Specifically, we model the association task as a similarity decoding problem and propose a sim-decoder with an association-centric learning mechanism. The sim-decoder calculates similarities in three aspects: spatial, temporal, and cross-clip. Subsequently, association-centric learning leverages these threefold similarities to ensure that the extracted features are appropriate for continuous tracking and robust enough to generalize to unknown categories. Compared with existing powerful OV-MOT and CV-MOT methods, AED achieves superior performance on TAO, SportsMOT, and DanceTrack without any prior knowledge. Our code is available at https://github.com/balabooooo/AED.

Text-to-Image Person Retrieval (TIPR) aims to retrieve person images based on natural language descriptions. Although many TIPR methods have achieved promising results, sometimes textual queries cannot accurately and comprehensively reflect the content of the image, leading to poor cross-modal alignment and overfitting to limited datasets. Moreover, the inherent modality gap between text and image further amplifies these issues, making accurate cross-modal retrieval even more challenging. To address these limitations, we propose the Generation-Enhanced Alignment (GEA) from a generative perspective. GEA contains two parallel modules: 1) Text-Guided Token Enhancement (TGTE), which introduces diffusion-generated images as intermediate semantic representations to bridge the gap between text and visual patterns. These generated images enrich the semantic representation of text and facilitate cross-modal alignment. 2) Generative Intermediate Fusion (GIF) module, which combines cross-attention between generated images, original images, and text features to generate a unified representation optimized by triplet alignment loss. We conduct extensive experiments on three public TIPR datasets, CUHK-PEDES, RSTPReid, and ICFG-PEDES, to evaluate the performance of GEA. The remarkable results justify the efficacy of our method. More implementation details and extended results are available at https://github.com/sugelamyd123/Sup-for-GEA.

Tracklet Association is an effective post-processing method for maintaining trajectory identity consistency in Multi-Object Tracking (MOT). Existing methods usually adopt two separate models to first split trajectories and then connect tracklets. However, they overlook the intrinsic relationship between the two tasks. In this work, we propose a graph neural network (GNN)-based framework to unify splitting and connection tasks in a single model that only needs to be trained once. Specifically, we summarize the commonalities of the two tasks by comparing the consistency of neighboring features and transform them into the edge prediction problems in GNN, i.e., where to split and whether to connect. We design two modules for global and local graph perception during message passing to effectively capture identity-consistent information of trajectories. Experiments conducted on the MOT17 and DanceTrack benchmarks demonstrate that our approach significantly improves IDF1 and HOTA metrics, showcasing its effectiveness in tracking consistently.

Single-stage multiport inverter (SSMI) is an attractive solution for hybrid electric vehicles. However, it is challenging to design the modulation scheme for the SSMI since flexible power split between sources on the dc side together with satisfactory stator current control on the ac side should be achieved simultaneously. To this end, a three-dimensional vector multimapping (3DVM)- based power split strategy is proposed in this article to realize flexible power split while guaranteeing the stator current regulation under unbalanced dc links. In the proposed 3DVM scheme, the asymmetrical 3-D vector space is divided into two symmetrical subspaces to simplify the modulation scheme design under unbalanced dc links, and the reference vector decomposed process provides an additional degree of freedom for power split by regulating the decomposed factors. Subsequently, a direct power control model with respect to decomposing factors is put forward to realize power split between dc-port sources without affecting the stator current control. Moreover, by mapping the decomposed reference vectors to the edges and planes of the subspaces, the switching actions are reduced, which increases the power conversion efficiency. Experiments are carried out to validate the effectiveness and superiority of the proposed scheme.

Joint Detection and Embedding(JDE) trackers have demonstrated excellent performance in Multi-Object Tracking(MOT) tasks by incorporating the extraction of appearance features as auxiliary tasks through embedding Re-Identification task(ReID) into the detector, achieving a balance between inference speed and tracking performance. However, solving the competition between the detector and the feature extractor has always been a challenge. Also, the issue of directly embedding the ReID task into MOT has remained unresolved. The lack of high discriminability in appearance features results in their limited utility. In this paper, we propose a new learning approach using cross-correlation to capture temporal information of objects. The feature extraction network is no longer trained solely on appearance features from each frame but learns richer motion features by utilizing feature heatmaps from consecutive frames, addressing the challenge of inter-class feature similarity. Furthermore, we apply our learning approach to a more lightweight feature extraction network, and treat the feature matching scores as strong cues rather than auxiliary cues, employing a appropriate weight calculation to reflect the compatibility between our obtained features and the MOT task. Our tracker, named TCBTrack, achieves state-of-the-art performance on multiple public benchmarks, i.e., MOT17, MOT20, and DanceTrack datasets. Specifically, on the DanceTrack test set, we achieve 56.8 HOTA, 58.1 IDF1 and 92.5 MOTA, making it the best online tracker that can achieve real-time performance. Comparative evaluations with other trackers prove that our tracker achieves the best balance between speed, robustness and accuracy.

Domain Generalizable Person Re-identification (DG ReID) is a practical task that generalizes the model trained on multiple source domains to unseen target domains without fine-tuning. Existing methods usually combine diverse normalization techniques to remove style information while retaining discriminative features. However, they overlook the absence of the target domain while training. In this work, we propose a Latent Distribution Alignment (LDA) method to indirectly align source and target domains through the dynamically constructed latent distribution without any added learnable parameters. Specifically, we design an expert network to dynamically construct a latent distribution and store the domain-specific representation information by fully utilizing batch-wise statistical parameters. Subsequently, one source domain, acting as the mimical target domain of the remaining source domains, explicitly aligns with the latent distribution through an instance-wise domain alignment network through Anti-Normalization (AN). Extensive experiments show that our method is simple yet effective in enhancing the generalizable capability.

Reconstruction-based methods have been commonly used for unsupervised anomaly detection, in which a normal image is reconstructed and compared with the given test image to detect and locate anomalies. Recently, diffusion models have shown promising applications for anomaly detection due to their powerful generative ability. However, these models lack strict mathematical support for normal image reconstruction and unexpectedly suffer from low reconstruction quality. To address these issues, this paper proposes a novel and highly-interpretable method named Masked Diffusion Posterior Sampling (MDPS). In MDPS, the problem of normal image reconstruction is mathematically modeled as multiple diffusion posterior sampling for normal images based on the devised masked noisy observation model and the diffusion-based normal image prior under Bayesian framework. Using a metric designed from pixel-level and perceptual-level perspectives, MDPS can effectively compute the difference map between each normal posterior sample and the given test image. Anomaly scores are obtained by averaging all difference maps for multiple posterior samples. Exhaustive experiments on MVTec and BTAD datasets demonstrate that MDPS can achieve state-of-the-art performance in normal image reconstruction quality as well as anomaly detection and localization.

Domain Generalizable Person Re-Identification (DG-ReID) strives to transfer learned feature representation from source domains to unseen target domains, despite significant distribution shifts. While most existing methods enhance model generalization and discriminative feature extraction capability by introducing Instance Normalization (IN) in combination with Batch Normalization (BN), these approaches still struggle with the overfitting of normalization layers to the source domains, posing challenges in domain generalization. To address this issue, we propose ReNorm, a purely normalization-based framework that integrates two complementary normalization layers through two forward propagations for the same weight matrix. In the first forward propagation, Remix Normalization (RN) combines IN and BN in a concise manner to ensure the feature extraction capability. As an effective complement to RN, Emulation Normalization (EN) simulates the testing process of RN, implicitly mitigating the domain shifts caused by the absence of target domain information and actively guiding the model in learning how to generalize the feature extraction capability to unseen target domains. Meanwhile, we propose Domain Frozen (DF), freezing updates to affine parameters to reduce the impact of these less robust parameters on overfitting to the source domains. Extensive experiments show that our framework achieves state-of-the-art performance on the popular benchmarks. The code is available at https://github.com/3699nr/ReNorm.

Link prediction for knowledge graphs (KGs), which aims to predict missing facts, has been broadly studied in binary relational KGs. However, real world data contains a large number of high-order interaction patterns, which is difficult to describe using only binary relations. In this work, we propose a relation-based dynamic learning model RD-MPNN, based on the message passing neural network model, to learn higher-order interactions and address the link prediction problem in knowledge hypergraphs. Different from existing methods, we consider the positional information of entities within a hyper-relation to differentiate each entity’s role in the hyper-relation. Furthermore, we complete the representation learning of hyper-relations by dynamically updating hyper-relations with entity information. Extensive evaluations on two representative knowledge hypergraph datasets demonstrate that our model outperforms the state-of-the-art methods. We also compare the performance of models at differing arities (the number of entities within a relation), to show that RD-MPNN demonstrates outstanding performance metrics for complex hypergraphs (arity>2).

The single-stage multiport inverter (SSMPI)-fed motor drives enable direct connection from the dc-side energy sources to the ac motor without utilizing any dc/dc converters, which feature high power density and high efficiency of the overall system. However, the output power of each source is highly coupled with the motor control in this topology, which provides poor flexibility of power control. To address this problem, a decoupling control strategy is proposed in this paper. First, an asymmetrical carrier-based modulation strategy is adopted to ensure the high performance of the motor control by fully considering the voltage variation of sources. Then, injecting different values of the zero-sequence voltage adds a degree of freedom for controlling the output power of sources. With the proposed method, the flexible power control is realized without affecting the performance of the motor control. Finally, the effectiveness of the proposed control scheme is verified through experiments on a permanent magnet synchronous motor (PMSM)-based laboratory prototype.

Person re-identification(ReID) aims to match and track people in a surveillance system with non-overlapping camera views. It is a key challenge for person ReID to learn robust and discriminative person representations. However, in the real world, similar person appearance, different image angles and changing person attributes make the task very difficult. To tackle this problem, we propose the attribute guided graph convolutional networks(AG-GCN) to design a model with stronger generalization. Specifically, an attribute transfer module is introduced into the framework to revise person attributes to obtain diverse person feature expression. In addition, we apply graph convolutional networks to combine attributes with body parts as a more fine-grained representation of the person. The experiment results conducted on Market 1501 and DukeMTMCReID datasets show that our method outperforms state-of-the-art attribute-based methods on a single dataset and generalizes better on other datasets.

Unsupervised domain adaptive (UDA) person re-identification (ReID) has gained increasing attention for its effectiveness on the target domain without manual annotations. Most fine-tuning based UDA person ReID methods focus on encoding global features for pseudo labels generation, neglecting the local feature that can provide for the fine-grained information. To handle this issue, we propose a Learning Feature Fusion (LF 2 ) framework for adaptively learning to fuse global and local features to obtain a more comprehensive fusion feature representation. Specifically, we first pre-train our model within a source domain, then fine-tune the model on unlabeled target domain based on the teacher-student training strategy. The average weighting teacher network is designed to encode global features, while the student network updating at each iteration is responsible for fine-grained local features. By fusing these multi-view features, multi-level clustering is adopted to generate diverse pseudo labels. In particular, a learnable Fusion Module (FM) for giving prominence to fine-grained local information within the global feature is also proposed to avoid obscure learning of multiple pseudo labels. Experiments show that our proposed LF 2 framework outperforms the state-of-the-art with 73.5% mAP and 83.7% Rank1 on Market1501 to DukeMTMC-ReID, and achieves 83.2% mAP and 92.8% Rank1 on DukeMTMC-ReID to Market1501. The source code is available at: https://github.com/DJEddyking/LF2.

Unsupervised domain adaptive (UDA) person re-identification (ReID) has gained increasing attention for its effectiveness on the target domain without manual annotations. Most fine-tuning based UDA person ReID methods focus on encoding global features for pseudo labels generation, neglecting the local feature that can provide for the fine-grained information. To handle this issue, we propose a Learning Feature Fusion (LF2) framework for adaptively learning to fuse global and local features to obtain a more comprehensive fusion feature representation. Specifically, we first pre-train our model within a source domain, then fine-tune the model on unlabeled target domain based on the teacher-student training strategy. The average weighting teacher network is designed to encode global features, while the student network updating at each iteration is responsible for fine-grained local features. By fusing these multi-view features, multi-level clustering is adopted to generate diverse pseudo labels. In particular, a learnable Fusion Module (FM) for giving prominence to fine-grained local information within the global feature is also proposed to avoid obscure learning of multiple pseudo labels. Experiments show that our proposed LF2 framework outperforms the state-of-the-art with 73.5% mAP and 83.7% Rank1 on Market1501 to DukeMTMC-ReID, and achieves 83.2% mAP and 92.8% Rank1 on DukeMTMC-ReID to Market1501.

Due to balanced accuracy and speed, one-shot models which jointly learn detection and identification embeddings, have drawn great attention in multi-object tracking (MOT). However, the inherent differences and relations between detection and re-identification (ReID) are unconsciously overlooked because of treating them as two isolated tasks in the one-shot tracking paradigm. This leads to inferior performance compared with existing two-stage methods. In this paper, we first dissect the reasoning process for these two tasks, which reveals that the competition between them inevitably would destroy task-dependent representations learning. To tackle this problem, we propose a novel reciprocal network (REN) with a self-relation and cross-relation design so that to impel each branch to better learn task-dependent representations. The proposed model aims to alleviate the deleterious tasks competition, meanwhile improve the cooperation between detection and ReID. Furthermore, we introduce a scale-aware attention network (SAAN) that prevents semantic level misalignment to improve the association capability of ID embeddings. By integrating the two delicately designed networks into a one-shot online MOT system, we construct a strong MOT tracker, namely CSTrack. Our tracker achieves the state-of-the-art performance on MOT16, MOT17 and MOT20 datasets, without other bells and whistles. Moreover, CSTrack is efficient and runs at 16.4 FPS on a single modern GPU, and its lightweight version even runs at 34.6 FPS. The complete code has been released at https://github.com/JudasDie/SOTS

To address the lack of accuracy and speed of aircraft detection in complex background images, the YOLOv5 model is improved in this paper to meet the growing demand for aircraft detection. Firstly, Mosaic-9 data enhancement is performed on the dataset, and MobileNet V3 Small is used to replace ResNet feature extraction network to enrich small target samples and improve feature extraction speed. Secondly, over-fitting is prevented by Label Smoothing, and channel pruning is used to reduce the complexity and redundancy of the neural network. Finally, TensorRT is used to optimize the underlying hardware resources to improve the detection speed. The experiments use Military Aircraft Detection Dataset for model training and validation on NVIDIA Tesla P100 16GB GPU. The results show that our improvements maintain the accuracy of the model while effectively increasing the training speed and detection speed of the model. Our codes are available at https://github.com/imcyx/Military-Detect