IJCAI 2023 论文列表

In this paper, we study how to achieve two characteristics highly-expected by incomplete multi-view clustering (IMvC). Namely, i) instance commonality refers to that within-cluster instances should share a common pattern, and ii) view versatility refers to that cross-view samples should own view-specific patterns. To this end, we design a novel dual-stream model which employs a dual attention layer and a dual contrastive learning loss to learn view-specific prototypes and model the sample-prototype relationship. When the view is missed, our model performs data recovery using the prototypes in the missing view and the sample-prototype relationship inherited from the observed view. Thanks to our dual-stream model, both cluster- and view-specific information could be captured, and thus the instance commonality and view versatility could be preserved to facilitate IMvC. Extensive experiments demonstrate the superiority of our method on five challenging benchmarks compared with 11 approaches. The code could be accessed from https://pengxi.me.

The increasing automation of high-stakes decisions with direct impact on the lives and well-being of individuals raises a number of important considerations. Prominent among these is strategic behavior by individuals hoping to achieve a more desirable outcome. Two forms of such behavior are commonly studied: 1) misreporting of individual attributes, and 2) recourse, or actions that truly change such attributes. The former involves deception, and is inherently undesirable, whereas the latter may well be a desirable goal insofar as it changes true individual qualification. We study misreporting and recourse as strategic choices by individuals within a unified framework. In particular, we propose auditing as a means to incentivize recourse actions over attribute manipulation, and characterize optimal audit policies for two types of principals, utility-maximizing and recourse-maximizing. Additionally, we consider subsidies as an incentive for recourse over manipulation, and show that even a utility-maximizing principal would be willing to devote a considerable amount of audit budget to providing such subsidies. Finally, we consider the problem of optimizing fines for failed audits, and bound the total cost incurred by the population as a result of audits.

Selecting influentials in networks against strategic manipulations has attracted many researchers' attention and it also has many practical applications. Here, we aim to select one or two influentials in terms of progeny (the influential power) and prevent agents from manipulating their edges (incentive compatibility). The existing studies mostly focused on selecting a single influential for this setting. Zhang et al. [2021] studied the problem of selecting one agent and proved an upper bound of 1/(1+ln2) to approximate the optimal selection. In this paper, we first design a mechanism to actually reach the bound. Then, we move this forward to choosing two agents and propose a mechanism to achieve an approximation ratio of (3+ln2)/(4(1+ln2)) (approx. 0.54).

A temporal graph has an edge set that may change over discrete time steps, and a temporal path (or walk) must traverse edges that appear at increasing time steps. Accordingly, two temporal paths (or walks) are temporally disjoint if they do not visit any vertex at the same time. The study of the computational complexity of finding temporally disjoint paths or walks in temporal graphs has recently been initiated by Klobas et al.. This problem is motivated by applications in multi-agent path finding (MAPF), which include robotics, warehouse management, aircraft management, and traffic routing. We extend Klobas et al.’s research by providing parameterized hardness results for very restricted cases, with a focus on structural parameters of the so-called underlying graph. On the positive side, we identify sufficiently simple cases where we can solve the problem efficiently. Our results reveal some surprising differences between the “path version” and the “walk version” (where vertices may be visited multiple times) of the problem, and answer several open questions posed by Klobas et al.

This study extends the recently-developed LaCAM algorithm for multi-agent pathfinding (MAPF). LaCAM is a sub-optimal search-based algorithm that uses lazy successor generation to dramatically reduce the planning effort. We present two enhancements. First, we propose its anytime version, called LaCAM*, which eventually converges to optima, provided that solution costs are accumulated transition costs. Second, we improve the successor generation to quickly obtain initial solutions. Exhaustive experiments demonstrate their utility. For instance, LaCAM* sub-optimally solved 99% of the instances retrieved from the MAPF benchmark, where the number of agents varied up to a thousand, within ten seconds on a standard desktop PC, while ensuring eventual convergence to optima; developing a new horizon of MAPF algorithms.

This paper studies multiparty learning, aiming to learn a model using the private data of different participants. Model reuse is a promising solution for multiparty learning, assuming that a local model has been trained for each party. Considering the potential sample selection bias among different parties, some heterogeneous model reuse approaches have been developed. However, although pre-trained local classifiers are utilized in these approaches, the characteristics of the local data are not well exploited. This motivates us to estimate the density of local data and design an auxiliary model together with the local classifiers for reuse. To address the scenarios where some local models are not well pre-trained, we further design a multiparty cross-entropy loss for calibration. Upon existing works, we address a challenging problem of heterogeneous model reuse from a decision theory perspective and take advantage of recent advances in density estimation. Experimental results on both synthetic and benchmark data demonstrate the superiority of the proposed method.

Allen's interval algebra is one of the most well-known calculi in qualitative temporal reasoning with numerous applications in artificial intelligence. Very recently, there has been a surge of improvements in the fine-grained complexity of NP-hard reasoning tasks in this algebra, which has improved the running time from the naive 2^O(n^2) to O*((1.0615n)^n), and even faster algorithms are known for unit intervals and the case when we a bounded number of overlapping intervals. Despite these improvements the best known lower bound is still only 2^o(n) under the exponential-time hypothesis and major improvements in either direction seemingly require fundamental advances in computational complexity. In this paper we propose a novel framework for solving NP-hard qualitative reasoning problems which we refer to as dynamic programming with sublinear partitioning. Using this technique we obtain a major improvement of O*((cn/log(n))^n) for Allen's interval algebra. To demonstrate that the technique is applicable to further problem domains we apply it to a problem in qualitative spatial reasoning, the cardinal direction calculus, and solve it in O*((cn/log(n))^(2n/3)) time. Hence, not only do we significantly advance the state-of-the-art for NP-hard qualitative reasoning problems, but obtain a novel algorithmic technique that is likely applicable to many problems where 2^O(n) time algorithms are unlikely.

Recent works reveal that adversarial augmentation benefits the generalization of neural networks (NNs) if used in an appropriate manner. In this paper, we introduce Temporal Adversarial Augmentation (TA), a novel video augmentation technique that utilizes temporal attention. Unlike conventional adversarial augmentation, TA is specifically designed to shift the attention distributions of neural networks with respect to video clips by maximizing a temporal-related loss function. We demonstrate that TA will obtain diverse temporal views, which significantly affect the focus of neural networks. Training with these examples remedies the flaw of unbalanced temporal information perception and enhances the ability to defend against temporal shifts, ultimately leading to better generalization. To leverage TA, we propose Temporal Video Adversarial Fine-tuning (TAF) framework for improving video representations. TAF is a model-agnostic, generic, and interpretability-friendly training strategy. We evaluate TAF with four powerful models (TSM, GST, TAM, and TPN) over three challenging temporal-related benchmarks (Something-something V1&V2 and diving48). Experimental results demonstrate that TAF effectively improves the test accuracy of these models with notable margins without introducing additional parameters or computational costs. As a byproduct, TAF also improves the robustness under out-of-distribution (OOD) settings. Code is available at https://github.com/jinhaoduan/TAF.

Imbalanced node classification widely exists in real-world networks where graph neural networks (GNNs) are usually highly inclined to majority classes and suffer from severe performance degradation on classifying minority class nodes. Various imbalanced node classification methods have been proposed recently which construct synthetic nodes and edges w.r.t. minority classes to balance the label/topology distribution. However, they are all based on homophilic assumption that nodes of the same label tend to connect despite the widely existence of heterophilic edges in real-world graphs. Thus, they uniformly aggregate features from both homophilic and heterophilic neighbors and rely on feature similarity to generate synthetic edges, which cannot be applied to imbalanced graphs in high heterophily. To address this problem, we propose a novel GraphSANN for imbalanced node classification on both homophilic and heterophilic graphs. Firstly, we propose a unified feature mixer to generate synthetic nodes with both homophilic and heterophilic interpolation in a unified way. Next, by randomly sampling edges between synthetic nodes and existing nodes as candidata edges, we design an adaptive subgraph extractor to dynamically extract the contextual subgraphs of candidate edges with flexible ranges. Finally, we develop a multi-filter subgraph encoder which constructs multiple different filter channels to discriminatively aggregate neighbors’ information along the homophilic and heterophilic edges. Extensive experiments on eight benchmark datasets demonstrate the superiority of our model for imbalanced node classificaiton on both homophilic and heterophilic graphs.

Handling occlusions is still a challenging problem for image composition. It always requires the source contents to be completely in front of the target contents or needs manual interventions to adjust occlusions, which is very tedious. Though several methods have suggested exploiting priors or learning techniques for promoting occlusion determination, their potentials are much limited. This paper addresses the challenge by presenting a depth registration method for merging the source contents seamlessly into the 3D space that the target image represents. Thus, the occlusions between the source contents and target contents can be conveniently handled through pixel-wise depth comparisons, allowing the user to more efficiently focus on the designs for image composition. Experimental results show that we can conveniently handle occlusions in image composition and improve efficiency by about 4 times compared to Photoshop.

Accurate completion of archaeological artifacts is a critical aspect in several archaeological studies, including documentation of variations in style, inference of chronological and ethnic groups, and trading routes trends, among many others. However, most available pottery is fragmented, leading to missing textural and morphological cues. Currently, the reassembly and completion of fragmented ceramics is a daunting and time-consuming task, done almost exclusively by hand, which requires the physical manipulation of the fragments. To overcome the challenges of manual reconstruction, reduce the materials' exposure and deterioration, and improve the quality of reconstructed samples, we present IberianVoxel, a novel 3D Autoencoder Generative Adversarial Network (3D AE-GAN) framework tested on an extensive database with complete and fragmented references. We generated a collection of 1001 3D voxelized samples and their fragmented references from Iberian wheel-made pottery profiles. The fragments generated are stratified into different size groups and across multiple pottery classes. Lastly, we provide quantitative and qualitative assessments to measure the quality of the reconstructed voxelized samples by our proposed method and archaeologists' evaluation.

The explosive popularity of smart manufacturing has caught the attention of researchers in terms of intelligent mold processing and management. Machining mold components is a crucial step in the mold production process for many industries, which creates (e.g., cutting, drilling, and shaping a metal) the individual parts (e.g., core pins, ejector pins, cavities, slides, and lifters) that make up a mold used in manufacturing. We present IMPsys, an AI-based system that automatically explores machining jobs, infers their processing time and schedules them on machines, given numerous 3D modelling files of mold components. Our demo video can be found at: http://bit.ly/3EeKnyL.

Knowledge distillation (KD) is an effective method for transferring the knowledge of a teacher model to a student model, that aims to improve the latter's performance efficiently. Although generic knowledge distillation methods such as softmax representation distillation and intermediate feature matching have demonstrated improvements with various tasks, only marginal improvements are shown in student networks due to their limited model capacity. In this work, to address the student model's limitation, we propose a novel flexible KD framework, Integrating Matched Features using Attentive Logit in Knowledge Distillation (IMF). Our approach introduces an intermediate feature distiller (IFD) to improve the overall performance of the student model by directly distilling the teacher's knowledge into branches of student models. The generated output of IFD, which is trained by the teacher model, is effectively combined by attentive logit. We use only a few blocks of the student and the trained IFD during inference, requiring an equal or less number of parameters. Through extensive experiments, we demonstrate that IMF consistently outperforms other state-of-the-art methods with a large margin over the various datasets in different tasks without extra computation.

Despite its success, generative adversarial networks (GANs) still suffer from mode collapse, i.e., the generator can only map latent variables to a partial set of modes in the target distribution. In this paper, we analyze and seek to regularize this issue with an independent and identically distributed (IID) sampling perspective and emphasize that holding the IID property referring to the target distribution for generation can naturally avoid mode collapse. This is based on the basic IID assumption for real data in machine learning. However, though the source samples {z} obey IID, the generations {G(z)} may not necessarily be IID sampling from the target distribution. Based on this observation, considering a necessary condition of IID generation, we propose a new loss to encourage the closeness between the inverse samples of real data and the Gaussian source in the latent space to regularize the generation to be IID from the target distribution. The logic is that the inverse samples from target data should also be IID in the source distribution. Experiments on both synthetic and real-world data show the effectiveness of our model.

The performance of nighttime semantic segmentation has been significantly improved thanks to recent unsupervised methods. However, these methods still suffer from complex domain gaps, i.e., the challenging illumination gap and the inherent dataset gap. In this paper, we propose the illumination-coupled domain adaptation framework(ICDA) to effectively avoid the illumination gap and mitigate the dataset gap by coupling daytime and nighttime images as a whole with semantic relevance. Specifically, we first design a new composite enhancement method(CEM) that considers not only illumination but also spatial consistency to construct the source and target domain pairs, which provides the basic adaptation unit for our ICDA. Next, to avoid the illumination gap, we devise the Deformable Attention Relevance(DAR) module to capture the semantic relevance inside each domain pair, which can couple the daytime and nighttime images at the feature level and adaptively guide the predictions of nighttime images. Besides, to mitigate the dataset gap and acquire domain-invariant semantic relevance, we propose the Prototype-based Class Alignment(PCA) module, which improves the usage of category information and performs fine-grained alignment. Extensive experiments show that our method reduces the complex domain gaps and achieves state-of-the-art performance for nighttime semantic segmentation. Our code is available at https://github.com/chenghaoDong666/ICDA.

Hyperspectral image (HSI) denoising has achieved promising results with the development of deep learning. A mainstream class of methods exploits the spatial-spectral correlations and recovers each band with the aids of neighboring bands, collectively referred to as spectral auxiliary networks. However, these methods treat entire adjacent spectral bands equally. In theory, clearer and nearer bands tend to contain more reliable spectral information than noisier and farther ones with higher uncertainties. How to achieve spectral enhancement and adaptation of each adjacent band has become an urgent problem in HSI denoising. This work presents the UA-Adjustor, a comprehensive adjustor that enhances denoising performance by considering both the band-to-pixel and enhancement-to-adjustment aspects. Specifically, UA-Adjustor consists of three stages that evaluate the importance of neighboring bands, enhance neighboring bands based on uncertainty perception, and adjust the weight of spatial pixels in adjacent bands through estimated uncertainty. For its simplicity, UA-Adjustor can be flexibly plugged into existing spectral auxiliary networks to improve denoising behavior at low cost. Extensive experimental results validate that the proposed solution can improve over recent state-of-the-art (SOTA) methods on both simulated and real-world benchmarks by a large margin.

Federated learning (FL) collaboratively models user data in a decentralized way. However, in the real world, non-identical and independent data distributions (non-IID) among clients hinder the performance of FL due to three issues, i.e., (1) the class statistics shifting, (2) the insufficient hierarchical information utilization, and (3) the inconsistency in aggregating clients. To address the above issues, we propose HyperFed which contains three main modules, i.e., hyperbolic prototype Tammes initialization (HPTI), hyperbolic prototype learning (HPL), and consistent aggregation (CA). Firstly, HPTI in the server constructs uniformly distributed and fixed class prototypes, and shares them with clients to match class statistics, further guiding consistent feature representation for local clients. Secondly, HPL in each client captures the hierarchical information in local data with the supervision of shared class prototypes in the hyperbolic model space. Additionally, CA in the server mitigates the impact of the inconsistent deviations from clients to server. Extensive studies of four datasets prove that HyperFed is effective in enhancing the performance of FL under the non-IID setting.

Creating a piece of music is difficult for people who have never been trained to compose. We present an automatic music generation system to lower the threshold of creating music. The system takes the user's humming as input and creates full music based on the humming melody. The system consists of five modules: 1) humming transcription, 2) melody generation, 3) broken chord generation, 4) accompaniment generation, and 5) audio synthesis. The first module transcribes the user's humming audio to a score, and then the melody generation module composes a complete melody based on the user's humming melody. After that, the third module will generate a broken chord track to accompany the full melody, and the fourth module will create more accompanying tracks. Finally, the audio synthesis module mixes all the tracks to generate the music. Through the user experiment, our system can generate high-quality music with natural expression based on the user's humming input.

My thesis consists of investigating how chest X-ray images, radiologists' eye movements and patients' clinical data can be used to teach a machine how radiologists read and classify images with the goal of creating human-centric AI architectures that can (1) capture radiologists' search behavioural patterns using their eye-movements in order to improve classification in DL systems, and (2) automatically detect lesions in medical images using clinical data and eye tracking data. Heterogeneous data sources such as chest X-rays, radiologists' eye movements, and patients' clinical data can contribute to novel multimodal DL architectures that, instead of learning directly from images' pixels, will learn human classification patterns encoded in both the eye movements of the images' regions and patients' medical history. In addition to a quantitative evaluation, I plan to conduct questionnaires with expert radiologists to understand the effectiveness of the proposed multimodal DL architecture.