Hard Instances Research Articles

Inverting Cryptographic Hash Functions via Cube-and-Conquer

MD4 and MD5 are fundamental cryptographic hash functions proposed in the early 1990s. MD4 consists of 48 steps and produces a 128-bit hash given a message of arbitrary finite size. MD5 is a more secure 64-step extension of MD4. Both MD4 and MD5 are vulnerable to practical collision attacks, yet it is still not realistic to invert them, i.e., to find a message given a hash. In 2007, the 39-step version of MD4 was inverted by reducing to SAT and applying a CDCL solver along with the so-called Dobbertin’s constraints. As for MD5, in 2012 its 28-step version was inverted via a CDCL solver for one specified hash without adding any extra constraints. In this study, Cube-and-Conquer (a combination of CDCL and lookahead) is applied to invert step-reduced versions of MD4 and MD5. For this purpose, two algorithms are proposed. The first one generates inverse problems for MD4 by gradually modifying the Dobbertin’s constraints. The second algorithm tries the cubing phase of Cube-and-Conquer with different cutoff thresholds to find the one with the minimum runtime estimate of the conquer phase. This algorithm operates in two modes: (i) estimating the hardness of a given propositional Boolean formula; (ii) incomplete SAT solving of a given satisfiable propositional Boolean formula. While the first algorithm is focused on inverting step-reduced MD4, the second one is not area-specific and is therefore applicable to a variety of classes of hard SAT instances. In this study, 40-, 41-, 42-, and 43-step MD4 are inverted for the first time via the first algorithm and the estimating mode of the second algorithm. Also, 28-step MD5 is inverted for four hashes via the incomplete SAT solving mode of the second algorithm. For three hashes out of them, it is done for the first time.

Multi-instance imbalance semantic segmentation by instance-dependent attention and adaptive hard instance mining

Multi-instance imbalanced semantic segmentation (MISS) is a common issue where larger foreground instances can dominate smaller ones, yet still produce satisfactory Dice. Detecting small instances is crucial for many applications. For example, in the diagnosis of coronary artery disease, it is essential to locate small-scale lesions. To address the challenge of MISS, we propose MISS-Net (MISS-Network), which considers the distribution characteristics and segmentation difficulty of instances. On one hand, given the spatially sparse distribution characteristics of instances, we propose an instance-dependent SparseViT. SparseViT learns unstructured sparse attention to capture instance-related patterns by computing content-aware attention scores and selecting the top-K indices. On the other hand, we propose a hard instance-aware loss that identifies ‘hard-to-segment’ instances through a dynamic instance weighting strategy. Combined with a global segmentation loss, it produces a trade-off between the global-level and instance-level segmentation performance. We extensively evaluated MISS-Net on three challenging 3D semantic segmentation tasks: two coronary artery calcification (CAC) datasets and one liver tumor (LITS) dataset. These datasets contain a large number of instances with significant variability in instance volumes (e.g., in the LITS dataset, a patient has an average of 12.39 instances, with instance volumes ranging from 987.66cm3 to 0.04cm3). The results demonstrate that MISS-Net significantly improves both instance-level and global-level segmentation performance. For the CAC task, MISS-Net improves Dice by 3.6% and 6.4%, and instance F1 by 3.1% and 1.9%. For the LITS dataset, Dice and instance F1 are improved by 1.6% and 3.8%, respectively.

Beyond quantum annealing: optimal control solutions to maxcut problems

Quantum Annealing (QA) relies on mixing two Hamiltonian terms, a simple driver and a complex problem Hamiltonian, in a linear combination. The time-dependent schedule for this mixing is often taken to be linear in time: improving on this linear choice is known to be essential and has proven to be difficult. Here, we present different techniques for improving on the linear-schedule QA along two directions, conceptually distinct but leading to similar outcomes: 1) the first approach consists of constructing a Trotter-digitized QA (dQA) with schedules parameterized in terms of Fourier modes or Chebyshev polynomials, inspired by the Chopped Random Basis algorithm for optimal control in continuous time; 2) the second approach is technically a Quantum Approximate Optimization Algorithm (QAOA), whose solutions are found iteratively using linear interpolation or expansion in Fourier modes. Both approaches emphasize finding smooth optimal schedule parameters, ultimately leading to hybrid quantum–classical variational algorithms of the alternating Hamiltonian Ansatz type. We apply these techniques to MaxCut problems on weighted 3-regular graphs with N = 14 sites, focusing on hard instances that exhibit a small spectral gap, for which a standard linear-schedule QA performs poorly. We characterize the physics behind the optimal protocols for both the dQA and QAOA approaches, discovering shortcuts to adiabaticity-like dynamics. Furthermore, we study the transferability of such smooth solutions among hard instances of MaxCut at different circuit depths. Finally, we show that the smoothness pattern of these protocols obtained in a digital setting enables us to adapt them to continuous-time evolution, contrarily to generic non-smooth solutions. This procedure results in an optimized QA schedule that is implementable on analog devices.

Generating hard quadratic unconstrained binary optimization instances via the method of combining bit reduction and duplication technique

ABSTRACT Quadratic Unconstrained Binary Optimization (QUBO) is a combinatorial optimization problem defined by an energy function that consists of a quadratic formula involving multiple binary variables. We propose the method of combing bit reduction and duplication technique that can generate hard instances of QUBO problems from any original QUBO problem without changing the size. The idea is to reduce the original QUBO problem for a specified number of bits and duplicate the same number of bits while ensuring all duplicated pair of bits take the same binary values. Experimental results show that generated QUBO instances are much hard for solving.

Scheduling for additive manufacturing with two-dimensional packing and incompatible items

Additive manufacturing technology can enable low-cost, efficient production of low-demand, highly-complex customized items with reduced lead time if production is judiciously planned. This paper addresses the assignment of requested items into batches and the scheduling of the batches onto 3D printers. The objective is to minimize the manufacturing makespan while satisfying items’ compatibility, two-dimensional no-overlap and containment packing constraints within a batch, and machine’s capability to manufacture a batch. The problem is modeled as a mixed integer linear program (MIP) that solves instances up to 100 items. For hard and large instances, this paper proposes a matheuristic that fathoms packings using a step-wise check procedure. Computational results reveal that the proposed heuristic improves the makespan of MIP solutions of hard instances by 12% on average, with improvements reaching up to 72% for instances with 150 items. They further show that the proposed heuristic finds the best makespan for 88% of all cases. Finally, we provide useful managerial insights for production flexibility and scheduling policies.

Finding and exploring promising search space for The 0–1 Multidimensional Knapsack Problem

The 0–1, Multidimensional Knapsack Problem (MKP) is a classical NP-hard combinatorial optimization problem with many engineering applications. In this paper, we propose a novel algorithm combining evolutionary computation with the exact algorithm to solve the 0–1 MKP. It maintains a set of solutions and utilizes the information from the population to extract good partial assignments. To find high-quality solutions, an exact algorithm is applied to explore the promising search space specified by the good partial assignments. The new solutions are used to update the population. Thus, the good partial assignments evolve towards a better direction with the improvement of the population. Extensive experimentation with commonly used benchmark sets shows that our algorithm outperforms the state-of-the-art heuristic algorithms, TPTEA and DQPSO, as well as the commercial solver CPlex. It finds better solutions than the existing algorithms and provides new lower bounds for 10 large and hard instances.

Streaming approximation resistance of every ordering CSP

An ordering constraint satisfaction problem (OCSP) is defined by a family F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal F$$\\end{document} of predicates mapping permutations on {1,…,k}\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\{1,\\ldots,k\\}$$\\end{document} to {0,1}\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\{0,1\\}$$\\end{document}. An instance of Max-OCSP(F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal F$$\\end{document}) on n variables consists of a list of constraints, each consisting of a predicate from F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal F$$\\end{document} applied on k distinct variables. The goal is to find an ordering of the n variables that maximizes the number of constraints for which the induced ordering on the k variables satisfies the predicate. OCSPs capture well-studied problems including ‘maximum acyclic subgraph’ (MAS) and “maximum betweenness”. In this work, we consider the task of approximating the maximum number of satisfiable constraints in the (single-pass) streaming setting, when an instance is presented as a stream of constraints. We show that for every F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal F$$\\end{document}, Max-OCSP(F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal F$$\\end{document}) is approximation-resistant to o(n)-space streaming algorithms, i.e., algorithms using o(n) space cannot distinguish streams where almost every constraint is satisfiable from streams where no ordering beats the random ordering by a noticeable amount. This space bound is tight up to polylogarithmic factors. In the case of MAS, our result shows that for every ϵ>0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\epsilon>0$$\\end{document}, MAS is not (1/2+ϵ)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$(1/2+\\epsilon)$$\\end{document}-approximable in o(n) space. The previous best inapproximability result, due to Guruswami & Tao (2019), only ruled out 3/4-approximations in o(n)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$o(\\sqrt n)$$\\end{document} space. Our results build on recent works of Chou et al. (2022b, 2024) who provide a tight, linear-space inapproximability theorem for a broad class of “standard” (i.e., non-ordering) constraint satisfaction problems (CSPs) over arbitrary (finite) alphabets. Our results are obtained by building a family of appropriate standard CSPs (one for every alphabet size q) from any given OCSP and applying their theorem to this family of CSPs. To convert the resulting hardness results for standard CSPs back to our OCSP, we show that the hard instances from this earlier theorem have the following “partition expansion” property with high probability: For every partition of the n variables into small blocks, for most of the constraints, all variables are in distinct blocks.

Pixelation of time matrices for solving permutation flowshop scheduling problem

Over the last decades, scheduling theory has been put into practice so as to tackle production process as a combinatorial optimisation problem and specifically as a PFSP (Permutation Flowshop Scheduling Problem). Often, NP-completeness of PFSP leads numerous research towards suggesting heuristic algorithms. The objective is to propose an optimal or near-to-optimal order of n jobs processing on m machines with a minimum completion time of all jobs. In this paper, a two-phase heuristic algorithm will be presented, named IFRS (Improved FRS). FRS, within the first phase, is going to be used to find a superior order of jobs out of Taillard’s instances; then the order of jobs will be improved again in phase 2. While running IFRS, a completely new idea, named Pixelation of time matrices, will be used for the very first time and provide a final pattern, which can be used in any heuristic algorithm with makespan criterion. After using the hard benchmark instances of Taillard, the superiority of IFRS over 20 algorithms is going to be shown by tables and statistical graphics. Due to its low makespan values, IFRS benefits flow production, as does NEH.

Artificial intelligence warm-start approach: optimizing the generalization capability of QAOA in complex energy landscapes

To address the issue of the quantum approximate optimization algorithm frequently encountering local minima and the cost of parameter optimization within complex non-convex optimization energy landscapes, we consider a warm-start method. This approach leverages the characteristics of transition states in the enhanced optimizer, specifically descending along unique negative curvature directions, to find smaller local minima. Our research results indicate that with the assistance of an enhanced pre-training structure of the AlphaZero AI model, the initialization generalization ability of the new optimizer is significantly enhanced across various test sets. We train on 2-SAT training sets with clause densities between α ≈ 2.6 and α ≈ 2.89, and transfer to more complex test sets. Additionally, the average residual energy density in transfer learning consistently remains below 0.01, even achieving a high transfer success probability of 98% in hard instances with α ≈ 3.7. The search efficiency, pre-trained by ensemble learning, was significantly enhanced, while only requiring simple interpolation of a few transition points to transfer on the global optimal solutions at higher sample clause densities.

Improving Oriented Object Detection by Scene Classification and Task-Aligned Focal Loss

Oriented object detection (OOD) can precisely detect objects with arbitrary direction in remote sensing images (RSIs). Up to now, the two-stage OOD methods have attracted more attention because of their high detection accuracy. However, the two-stage methods only rely on the features of each proposal for object recognition, which leads to the misclassification problem because of the intra-class diversity, inter-class similarity and clutter backgrounds in RSIs. To address the above problem, an OOD model combining scene classification is proposed. Considering the fact that each foreground object has a strong contextual relationship with the scene of the RSI, a scene classification branch is added to the baseline OOD model, and the scene classification result of input RSI is used to exclude the impossible categories. To focus on the hard instances and enhance the consistency between classification and regression, a task-aligned focal loss (TFL) which combines the classification difficulty with the regression loss is proposed, and TFL assigns lager weights to the hard instances and optimizes the classification and regression branches simultaneously. The ablation study proves the effectiveness of scene classification branch, TFL and their combination. The comparisons with 15 and 14 OOD methods on the DOTA and DIOR-R datasets validate the superiority of our method.

Evolutionary optimization approach based on heuristic information with pseudo-utility for the quadratic assignment problem

Existing data mining-based meta-heuristic approaches have endeavored to efficiently solve the Quadratic Assignment Problem (QAP). However, their captured heuristic information lacks com prehensiveness and usefulness, as they fail to consider various combinations of frequent items and ignore objective information. This deficiency makes it challenging to generate new solutions that balance global and local search. To address these issues, we propose an approach called EOHIPU. EOHIPU efficiently mines and represents more itemsets composed of frequent items, and simultaneously introduces pseudo-utility to quantitatively characterize each item's contribution to the objective, thereby capturing more comprehensive and useful heuristic information. With this heuristic information, EOHIPU designs a mechanism for generating new solutions, balancing global and local search. We conduct experiments on 21 hard benchmark instances from QAPLIB, comparing EOHIPU with five state-of-the-art approaches. Our EOHIPU achieves the best results, with average percentage deviations (APDs) of 0.064 and 0.046 under short and long running times, respectively. Additionally, EOHIPU demonstrates faster convergence compared to other state-of-the-art approaches across all instances. These results highlight EOHIPU's superior solution quality and convergence rate. Moreover, the effectiveness of the captured heuristic information and the new solution generation mechanism is validated.

Algorithms for the ferromagnetic Potts model on expanders

AbstractWe give algorithms for approximating the partition function of the ferromagnetic $q$ -color Potts model on graphs of maximum degree $d$ . Our primary contribution is a fully polynomial-time approximation scheme for $d$ -regular graphs with an expansion condition at low temperatures (that is, bounded away from the order-disorder threshold). The expansion condition is much weaker than in previous works; for example, the expansion exhibited by the hypercube suffices. The main improvements come from a significantly sharper analysis of standard polymer models; we use extremal graph theory and applications of Karger’s algorithm to count cuts that may be of independent interest. It is #BIS-hard to approximate the partition function at low temperatures on bounded-degree graphs, so our algorithm can be seen as evidence that hard instances of #BIS are rare. We also obtain efficient algorithms in the Gibbs uniqueness region for bounded-degree graphs. While our high-temperature proof follows more standard polymer model analysis, our result holds in the largest-known range of parameters $d$ and $q$ .

Generative-Based Fusion Mechanism for Multi-Modal Tracking

Generative models (GMs) have received increasing research interest for their remarkable capacity to achieve comprehensive understanding. However, their potential application in the domain of multi-modal tracking has remained unexplored. In this context, we seek to uncover the potential of harnessing generative techniques to address the critical challenge, information fusion, in multi-modal tracking. In this paper, we delve into two prominent GM techniques, namely, Conditional Generative Adversarial Networks (CGANs) and Diffusion Models (DMs). Different from the standard fusion process where the features from each modality are directly fed into the fusion block, we combine these multi-modal features with random noise in the GM framework, effectively transforming the original training samples into harder instances. This design excels at extracting discriminative clues from the features, enhancing the ultimate tracking performance. Based on this, we conduct extensive experiments across two multi-modal tracking tasks, three baseline methods, and four challenging benchmarks. The experimental results demonstrate that the proposed generative-based fusion mechanism achieves state-of-the-art performance by setting new records on GTOT, LasHeR and RGBD1K. Code will be available at https://github.com/Zhangyong-Tang/GMMT.

Learning in Online Principal-Agent Interactions: The Power of Menus

We study a ubiquitous learning challenge in online principal-agent problems during which the principal learns the agent's private information from the agent's revealed preferences in historical interactions. This paradigm includes important special cases such as pricing and contract design, which have been widely studied in recent literature. However, existing work considers the case where the principal can only choose a single strategy at every round to interact with the agent and then observe the agent's revealed preference through their actions. In this paper, we extend this line of study to allow the principal to offer a menu of strategies to the agent and learn additionally from observing the agent's selection from the menu. We provide a thorough investigation of several online principal-agent problem settings and characterize their sample complexities, accompanied by the corresponding algorithms we have developed. We instantiate this paradigm to several important design problems — including Stackelberg (security) games, contract design, and information design. Finally, we also explore the connection between our findings and existing results about online learning in Stackelberg games, and we offer a solution that can overcome a key hard instance of previous work.

A Novel Evolutionary Algorithm With Column and Sub-Block Local Search for Sudoku Puzzles

Sudoku puzzles are not only popular intellectual games but also NP-hard combinatorial problems related to various real-world applications, which have attracted much attention worldwide. Although many efficient tools, such as evolutionary computation (EC) algorithms, have been proposed for solving Sudoku puzzles, they still face great challenges with regard to hard and large instances of Sudoku puzzles. Therefore, to efficiently solve Sudoku puzzles, this paper proposes a genetic algorithm (GA)-based method with a novel local search technology called local search-based GA (LSGA). The LSGA includes three novel design aspects. First, it adopts a matrix coding scheme to represent individuals and designs the corresponding crossover and mutation operations. Second, a novel local search strategy based on column search and sub-block search is proposed to increase the convergence speed of the GA. Third, an elite population learning mechanism is proposed to let the population evolve by learning the historical optimal solution. Based on the above technologies, LSGA can greatly improve the search ability for solving complex Sudoku puzzles. LSGA is compared with some state-of-the-art algorithms at Sudoku puzzles of different difficulty levels and the results show that LSGA performs well in terms of both convergence speed and success rates on the tested Sudoku puzzle instances.

Exploiting symmetry for the job sequencing and tool switching problem

The Job Sequencing and Tool Switching Problem (SSP), a well-known combinatorial optimization problem in the domain of Flexible Manufacturing Systems (FMS), is studied in this article. The aim of this research is to improve the currently known exact solution methodology for this NP-hard problem. We propose a new integer linear programming approach with symmetry-breaking and tightening cuts that provably outperformed the existing methodology described in the literature. The computational study conducted on a data set with 1,050 instances showed that the developed method can solve hard instances that were not solvable with the state-of-the-art methods presented in the literature.

Enhanced methods for the weight constrained shortest path problem

AbstractThe classic problem of constrained pathfinding is a well‐studied, yet challenging, network optimization problem with a broad range of applications in various areas such as communication and transportation. The weight constrained shortest path problem (WCSPP), the base form of constrained pathfinding with only one side constraint, aims to plan a cost‐optimum path with limited weight/resource usage. Given the bi‐criteria nature of the problem (i.e., dealing with the cost and weight of paths), methods addressing the WCSPP have some common properties with bi‐objective search. This article leverages the recent state‐of‐the‐art techniques in both constrained pathfinding and bi‐objective search and presents two new solution approaches to the WCSPP on the basis of A* search, both capable of solving hard WCSPP instances on very large graphs. We empirically evaluate the performance of our algorithms on a set of large and realistic problem instances and show their advantages over the state‐of‐the‐art algorithms in both time and space metrics. This article also investigates the importance of priority queues in constrained search with A*. We show with extensive experiments on both realistic and randomized graphs how bucket‐based queues without tie‐breaking can effectively improve the algorithmic performance of exhaustive A*‐based bi‐criteria searches.

An Attention-Based Method for the Minimum Vertex Cover Problem on Complex Networks

Solving combinatorial problems on complex networks represents a primary issue which, on a large scale, requires the use of heuristics and approximate algorithms. Recently, neural methods have been proposed in this context to find feasible solutions for relevant computational problems over graphs. However, such methods have some drawbacks: (1) they use the same neural architecture for different combinatorial problems without introducing customizations that reflects the specificity of each problem; (2) they only use a nodes local information to compute the solution; (3) they do not take advantage of common heuristics or exact algorithms. Following this interest, in this research we address these three main points by designing a customized attention-based mechanism that uses both local and global information from the adjacency matrix to find approximate solutions for the Minimum Vertex Cover Problem. We evaluate our proposal with respect to a fast two-factor approximation algorithm and a widely adopted state-of-the-art heuristic both on synthetically generated instances and on benchmark graphs with different scales. Experimental results demonstrate that, on the one hand, the proposed methodology is able to outperform both the two-factor approximation algorithm and the heuristic on the test datasets, scaling even better than the heuristic with harder instances and, on the other hand, is able to provide a representation of the nodes which reflects the combinatorial structure of the problem.

Target-aware transformer tracking with hard occlusion instance generation.

Visual tracking is a crucial task in computer vision that has been applied in diverse fields. Recently, transformer architecture has been widely applied in visual tracking and has become a mainstream framework instead of the Siamese structure. Although transformer-based trackers have demonstrated remarkable accuracy in general circumstances, their performance in occluded scenes remains unsatisfactory. This is primarily due to their inability to recognize incomplete target appearance information when the target is occluded. To address this issue, we propose a novel transformer tracking approach referred to as TATT, which integrates a target-aware transformer network and a hard occlusion instance generation module. The target-aware transformer network utilizes an encoder-decoder structure to facilitate interaction between template and search features, extracting target information in the template feature to enhance the unoccluded parts of the target in the search features. It can directly predict the boundary between the target region and the background to generate tracking results. The hard occlusion instance generation module employs multiple image similarity calculation methods to select an image pitch in video sequences that is most similar to the target and generate an occlusion instance mimicking real scenes without adding an extra network. Experiments on five benchmarks, including LaSOT, TrackingNet, Got10k, OTB100, and UAV123, demonstrate that our tracker achieves promising performance while running at approximately 41 fps on GPU. Specifically, our tracker achieves the highest AUC scores of 65.5 and 61.2% in partial and full occlusion evaluations on LaSOT, respectively.

Discriminative Identity-Feature Exploring and Differential Aware Learning for Unsupervised Person Re-Identification

Unsupervised person re-identification (Re-ID) aims to learn discriminative representations for person retrieval from unlabeled data. Currently, state-of-the-art techniques accomplish this task by using instance contrastive learning, which contrasts the similarities of the instances in different views. However, existing contrastive methods only focus on the positive effects of inter-instance relationships, while neglecting the negative effects of intra-instance redundancy information. This redundancy information can generate invalid or spurious intra-class relationships during the instance contrasting process, which enlarges the intra-class gaps and increases the noisy pseudo-labels. To address this issue, we propose a discriminative identity-feature exploring and differential aware learning (DiDAL) framework to learn more discriminative intra-identity representations. Specifically, the DiDAL extracts intra-instance salient features by synthetic complementary attention, and further explores the discriminative identity features by modeling the relationship among these salient features based on graph neural networks. This strategy aims to reduce the intra-instance redundancy information. Moreover, DiDAL explores hard instances by leveraging the extracted intra-instance salient features, and matches an anchor with multiple hard positive instances to enhance the robustness of the model to noisy pseudo-labels. Extensive experiment results on two widely used person re-identification datasets and a vehicle re-identification dataset demonstrate the superiority of the proposed method compared with existing state-of-the-art methods.