IJCAI 2024 论文列表

Machine learning based molecular property prediction has been a hot topic in the field of computer aided drug discovery (CADD). However, current MPP methods face two prominent challenges: 1) single-view MPP methods do not sufficiently exploit the complementary information of molecular data across multiple views, generally producing suboptimal performance, and 2) most existing multi-view MPP methods ignore the disparities in data quality among different views, inadvertently introducing the risk of models being overshadowed by inferior views. To address the above challenges, we introduce a novel cross-view contrastive fusion for enhanced molecular property prediction method (MolFuse). First, we extract intricate molecular semantics and structures from both sequence and graph views to leverage the complementarity of multi-view data. Then, MolFuse employs two distinct graphs, the atomic graph and chemical bond graph, to enhance the representation of the molecular graph, allow us to integrate both the fundamental backbone attributes and the nuanced shape characteristics. Notably, we incorporate a dual learning mechanism to refine the initial feature representations, and global features are obtained by maximizing the coherence among diverse view-specific molecular representations for the downstream task. The overall learning processes are combined into a unified optimization problem for iterative training. Experiments on multiple benchmark datasets demonstrate the superiority of our MolFuse.

While far-field multi-talker mixtures are recorded, each speaker can wear a close-talk microphone so that close-talk mixtures can be recorded at the same time. Although each close-talk mixture has a high signal-to-noise ratio (SNR) of the wearer, it has a very limited range of applications, as it also contains significant cross-talk speech by other speakers and is not clean enough. In this context, we propose a novel task named \textit{cross-talk reduction} (CTR) which aims at reducing cross-talk speech, and a novel solution named CTRnet which is based on unsupervised or weakly-supervised neural speech separation. In unsupervised CTRnet, close-talk and far-field mixtures are stacked as input for a DNN to estimate the close-talk speech of each speaker. It is trained in an unsupervised, discriminative way such that the DNN estimate for each speaker can be linearly filtered to cancel out the speaker's cross-talk speech captured at other microphones. In weakly-supervised CTRnet, we assume the availability of each speaker's activity timestamps during training, and leverage them to improve the training of unsupervised CTRnet. Evaluation results on a simulated two-speaker CTR task and on a real-recorded conversational speech separation and recognition task show the effectiveness and potential of CTRnet.

Deep learning-based pansharpening methods typically use simulated data at the reduced-resolution scale for training. It limits their performance when generalizing the trained model to the full-resolution scale due to incomprehensive information utilization of panchromatic (PAN) images at the full-resolution scale and low generalization ability. In this paper, we adopt two targeted strategies to address the above two problems. On the one hand, we introduce a cross-scale comprehensive information capture module, which improves the information utilization of the original PAN image through fully-supervised reconstruction. On the other hand, we pioneer a domain adaptation strategy to tackle the problem of low generalization across different scales. Considering the instinct domain gap between different scales, we leverage the maximum mean discrepancy loss and the inherent pixel-level correlations between features at different scales to reduce the scale variance, thus boosting the generalization ability of our model. Experiments on various satellites demonstrate the superiority of our method over the state-of-the-arts in terms of information retention. Our code is publicly available at https://github.com/Meiqi-Gong/SDIPS.

Existing neural heuristics often train a deep architecture from scratch for each specific vehicle routing problem (VRP), ignoring the transferable knowledge across different VRP variants. This paper proposes the cross-problem learning to assist heuristics training for different downstream VRP variants. Particularly, we modularize neural architectures for complex VRPs into 1) the backbone Transformer for tackling the travelling salesman problem (TSP), and 2) the additional lightweight modules for processing problem-specific features in complex VRPs. Accordingly, we propose to pre-train the backbone Transformer for TSP, and then apply it in the process of fine-tuning the Transformer models for each target VRP variant. On the one hand, we fully fine-tune the trained backbone Transformer and problem-specific modules simultaneously. On the other hand, we only fine-tune small adapter networks along with the modules, keeping the backbone Transformer still. Extensive experiments on typical VRPs substantiate that 1) the full fine-tuning achieves significantly better performance than the one trained from scratch, and 2) the adapter-based fine-tuning also delivers comparable performance while being notably parameter-efficient. Furthermore, we empirically demonstrate the favorable effect of our method in terms of cross-distribution application and versatility.

Cross-Domain Few-shot Semantic Segmentation (CD-FSS) aims to train generalized models that can segment classes from different domains with a few labeled images. Previous works have proven the effectiveness of feature transformation in addressing CD-FSS. However, they completely rely on support images for feature transformation, and repeatedly utilizing a few support images for each class may easily lead to overfitting and overlooking intra-class appearance differences. In this paper, we propose a Doubly Matching Transformation-based Network (DMTNet) to solve the above issue. Instead of completely relying on support images, we propose Self-Matching Transformation (SMT) to construct query-specific transformation matrices based on query images themselves to transform domain-specific query features into domain-agnostic ones. Calculating query-specific transformation matrices can prevent overfitting, especially for the meta-testing stage where only one or several images are used as support images to segment hundreds or thousands of images. After obtaining domain-agnostic features, we exploit a Dual Hypercorrelation Construction (DHC) module to explore the hypercorrelations between the query image with the foreground and background of the support image, based on which foreground and background prediction maps are generated and supervised, respectively, to enhance the segmentation result. In addition, we propose a Test-time Self-Finetuning (TSF) strategy to more accurately self-tune the query prediction in unseen domains. Extensive experiments on four popular datasets show that DMTNet achieves superior performance over state-of-the-art approaches. Code is available at https://github.com/ChenJiayi68/DMTNet.

Domain generalization aims to develop models that are robust to distribution shifts. Existing methods focus on learning invariance across domains to enhance model robustness, and data augmentation has been widely used to learn invariant predictors, with most methods performing augmentation in the input space. However, augmentation in the input space has limited diversity whereas in the feature space is more versatile and has shown promising results. Nonetheless, feature semantics is seldom considered and existing feature augmentation methods suffer from a limited variety of augmented features. We decompose features into class-generic, class-specific, domain-generic, and domain-specific components. We propose a cross-domain feature augmentation method named XDomainMix that enables us to increase sample diversity while emphasizing the learning of invariant representations to achieve domain generalization. Experiments on widely used benchmark datasets demonstrate that our proposed method is able to achieve state-of-the-art performance. Quantitative analysis indicates that our feature augmentation approach facilitates the learning of effective models that are invariant across different domains.

In this paper we study the Hospitals/Residents problem with Couples (HRC), where a solution is a stable matching or a report that none exists. We present a novel polynomial-time algorithm that can find a near-feasible stable matching (adjusting the hospitals' capacities by at most 1) in an HRC instance where the couples' preferences are sub-responsive (i.e., if one member switches to a better hospital, than the couple also improves if the new pair is also acceptable) and sub-complete (i.e., each pair of hospitals that are individually acceptable to both members are jointly acceptable for the couple) by reducing it to an instance of the Stable Fixtures problem. We also present a polynomial-time algorithm for HRC in a sub-responsive, sub-complete instance that is a Dual Market, or where all couples are one of several possible types. Our polynomial-time solvability results greatly expand the class of known tractable instances of HRC. We complement our algorithms with several hardness results. We show that HRC with sub-responsive and sub-complete couples is NP-hard, even with other strong restrictions. We also show that HRC with a Dual Market is NP-hard under several simultaneous restrictions.

With the proliferation of Location-based Social Networks (LBSNs), user check-in data at Points-of-Interest (POIs) has surged, offering rich insights into user preferences. However, sequential POI recommendation systems always face two pivotal challenges. A challenge lies in the difficulty of modeling time in a discrete space, which fails to accurately capture the dynamic nature of user preferences. Another challenge is the inherent sparsity and noise in continuous POI recommendation, which hinder the recommendation process. To address these challenges, we propose counterfactual user sequence synthesis with continuous time dynamic preference modeling (CussCtpm). CussCtpm innovatively combines Gated Recurrent Unit (GRU) with neural Ordinary Differential Equations (ODEs) to model user preferences in a continuous time framework. CussCtpm captures user preferences at both the POI-level and interest-level, identifying deterministic and non-deterministic preference concepts. Particularly at the interest-level, we employ GRU and neural ODEs to model users' dynamic preferences in continuous space, aiming to capture finer-grained shifts in user preferences over time. Furthermore, CussCtpm utilizes counterfactual data augmentation to generate counterfactual positive and negative user sequences. Our extensive experiments on two widely-used public datasets demonstrate that CussCtpm outperforms several advanced baseline models.

We propose regular expression inference (REI) as a challenge for code/language modelling, and the wider machine learning community. REI is a supervised machine learning (ML) and program optimisation task, and poses the problem of finding minimal regular expressions from examples: Given two finite sets of strings P and N and a cost function cost(·), the task is to generate an expression r that accepts all strings in P and rejects all strings in N , while no other such expression r′ exists with cost(r′) < cost(r). REI has advantages as a challenge problem: (i) regular expressions are well-known, widely used, and a natural idealisation of code; (ii) REI’s asymptotic worst-case complexity is well understood; (iii) REI has a small number of easy to understand parameters (e.g. P or N cardinality, string lengths of examples, or the cost function); this lets us easily finetune REI-hardness; (iv) REI, with its emphasis on optimisation, is an unsolved problem for deep learning based ML. Recently, an REI solver was implemented on GPUs, using program synthesis techniques. This enabled, for the first time, fast generation of minimal regular expressions for complex REI instances. Building on this advance, we generate and publish the first large-scale datasets for REI, and devise and evaluate several initial heuristic and machine learning baselines. We invite the community to participate and explore ML methods that learn to solve REI problems. We believe that progress in REI directly translates to progress in code/language modelling.

The multi-modal classification methods based on neural architecture search (NAS-MMC) can automatically learn a satisfied classifier from a given multi-modal search space. However, as the number of multi-modal features and fusion operators increases, the complexity of search space has increased dramatically. Rapidly identifying the satisfied fusion model from this vast space is very challenging. In this paper, we propose an efficient NAS-MMC method based on an idea of shrink-and-expansion search space, called core-structure-guided neural architecture search (CSG-NAS). Specifically, an evolutionary algorithm is first used to find core structures from a shrunk space (also called core structure search space) determined by high-quality features and fusion operators. Then a local search algorithm is used to find the optimal MMC model from the expanded space determined by the discovered core structures and the rest features as well as fusion operators. Moreover, a knowledge transfer strategy is introduced to further improve the overall performance and efficiency of the entire search process. Finally, extensive experimental results demonstrate the effectiveness of our CSG-NAS, attaining the superiority of classification performance, training efficiency and model complexity, compared to state-of-the-art ompetitors on several public benchmark multi-modal tasks. The source code is available at https://github.com/fupinhan123/CSG-NAS.

Across disciplines, cooperation is a fundamental research topic. While socially desirable to a population, it often bears a cost to the individual who, in their own self-interest, rationally chooses not to engage in costly cooperation. As such, much work has been done in understanding the biological mechanisms behind cooperation in human and animal populations. In my PhD project, I develop and apply these mechanisms both to artificial multi-agent systems and real social systems. I examine how factors such as agent heterogeneity and different learning algorithms affect not only the level of cooperation within a system, but also the level of fairness in the distribution of payoffs. In previous work, I showed how the effectiveness of the social norm-based mechanism of indirect reciprocity is affected when in-group biased cooperation is present. Beyond my future work on online platforms, I also plan to explore the effects of space, gossip, and partial and subjective observations to widen the potential scope of applications.

Many settings of interest involving humans and machines – from virtual personal assistants to autonomous vehicles – can naturally be modelled as principals (humans) delegating to agents (machines), which then interact with each other on their principals’ behalf. We refer to these multi-principal, multi-agent scenarios as delegation games. In such games, there are two important failure modes: problems of control (where an agent fails to act in line their principal’s preferences) and problems of cooperation (where the agents fail to work well together). In this paper we formalise and analyse these problems, further breaking them down into issues of alignment (do the players have similar preferences?) and capabilities (how competent are the players at satisfying those preferences?). We show – theoretically and empirically – how these measures determine the principals’ welfare, how they can be estimated using limited observations, and thus how they might be used to help us design more aligned and cooperative AI systems.

Knowing that a function is convex ensures that any local minimum is also a global minimum. Here, we implement an approach to certify the convexity of twice-differentiable functions by certifying that their second-order derivative is positive semidefinite. Both the computation of the second-order derivative and the certification of positive semidefiniteness are done symbolically. Previous implementations of this approach assume that the function to be minimized takes scalar or vector inputs, meaning that the second-order derivative is at most a matrix. However, the input of many machine learning problems is naturally given in the form of matrices or higher order tensors, in which case the second-order derivative becomes a tensor of at least fourth order. The familiar linear algebra notations and known rules for determining whether a matrix is positive semidefinite are not sufficient to deal with these higher order expressions. Here, we present a formal language for tensor expressions that allows us to generalize semidefiniteness to higher-order tensors and thereby certify the convexity of a broader set of functions.

Existing Deep Multi-view Clustering (DMVC) approaches typically concentrate on capturing consensus semantics from multiple views, where contrastive learning is widely used to align view-specific representations of each view. Unfortunately, view-specific representations are extracted from the content information of the corresponding instance, neglecting the relationships among different instances. Furthermore, existing contrastive loss imports numerous false negative pairs that conflict with the clustering objectives. In response to these challenges, we propose a contraStive and viEw-interaction stRucture learning framework for multI-viEw cluStering (SERIES). Our method takes into account the structural relations among instances and boosts the contrastive loss to improve intra-class compactness. Meanwhile, a cross-view dual relation generation mechanism is introduced to achieve the consensus structural graph across multiple views for clustering. Specifically, we initially acquire view-specific representations using multiple graph autoencoders to exploit both content information and structural information. Furthermore, to pull together the same cluster instances, a soft negative pair aware contrastive loss is employed to distinguish the dissimilar instances while attracting similar instances. Thereafter, the view-specific representations are fed into cross-view dual relation generation layers to generate the affinity matrices of each other, aiming to reveal a consistent structural graph across various views. Extensive experiments conducted on six benchmarks illustrate the superiority of our method compared to other state-of-the-art approaches.

Hashing utilizes hash code as a compact image representation, offering excellent performance in large-scale image retrieval due to its computational and storage advantages. However, the prevalence of degraded images on social media platforms, resulting from imperfections in the image capture process, poses new challenges for conventional image retrieval methods. To address this issue, we propose Contrastive Transformer Masked Image Hashing (CTMIH), a novel deep unsupervised hashing method specifically designed for degraded image retrieval, which is challenging yet relatively less studied. CTMIH addresses the problem by training on transformed and masked images, aiming to learn transform-invariant hash code in an unsupervised manner to mitigate performance degradation caused by image deterioration. CTMIH utilizes Vision Transformer (ViT) architecture applied to image patches to capture distant semantic relevance. CTMIH introduces cross-view debiased contrastive loss to align hash tokens from augmented views of the same image and presents semantic mask reconstruction loss at the patch level to recover masked patch tokens. Extensive empirical studies conducted on three benchmark datasets demonstrate the superiority of the proposed CTMIH over the state-of-the-art in both degraded and normal image retrieval.

As video-based social networks continue to grow exponentially, there is a rising interest in video retrieval using natural language. Cross-modal hashing, which learns compact hash code for encoding multi-modal data, has proven to be widely effective in large-scale cross-modal retrieval, e.g., image-text retrieval, primarily due to its computation and storage efficiency. However, when applied to video-text retrieval, existing cross-modal hashing methods generally extract features at the frame- or word-level for videos and texts individually, thereby ignoring their long-term dependencies. To address this issue, we propose Contrastive Transformer Cross-Modal Hashing (CTCH), a novel approach designed for video-text retrieval task. CTCH employs bidirectional transformer encoder to encode video and text and leverages their long-term dependencies. CTCH further introduces supervised multi-modality contrastive loss that effectively exploits inter-modality and intra-modality similarities among videos and texts. The experimental results on three video benchmark datasets demonstrate that CTCH outperforms the state-of-the-arts in video-text retrieval tasks.

Taxonomy completion, a self-supervised task, aims to add new concepts to an existing taxonomy by attaching them to appropriate hypernym and hyponym pairs. Researchers have proposed several approaches to capture the essential relationships in taxonomy using semantic or structural information. However, they either construct training signals from a single view or simply use a random sampling strategy, making it insufficient to capture various relations in taxonomic structure and learn quality representations. To address this, we propose CoSTC, a contrastive learning framework that captures diverse relations and improves representations for taxonomy completion. It uses two contrasting views, namely intra-view and inter-view, to provide rich self-supervised signals. In intra-view contrasting, we exploit the correlations within queries and within positions by performing instance-level discrimination task. In inter-view contrasting, we use a sampling strategy that considers diversity and hardness to select representative pairs, enhancing the learning of fine-grained query-position relations. Experimental results on three datasets verify the effectiveness of our approach. Our code is available at https://github.com/nyh-a/CoSTC.

Despite recent advancements in Self-Supervised Learning (SSL) for Time Series analysis, a noticeable gap persists between the anticipated achievements and actual performance. While these methods have demonstrated formidable generalization capabilities with minimal labels in various domains, their effectiveness in distinguishing between different classes based on a limited number of annotated records is notably lacking. Our hypothesis attributes this bottleneck to the prevalent use of Contrastive Learning, a shared training objective in previous state-of-the-art (SOTA) methods. By mandating distinctiveness between representations for negative pairs drawn from separate records, this approach compels the model to encode unique record-based patterns but simultaneously neglects changes occurring across the entire record. To overcome this challenge, we introduce Distilled Embedding for Almost-Periodic Time Series (DEAPS) in this paper, offering a non-contrastive method tailored for quasiperiodic time series, such as electrocardiogram (ECG) data. By avoiding the use of negative pairs, we not only mitigate the model's blindness to temporal changes but also enable the integration of a "Gradual Loss (L_gra)" function. This function guides the model to effectively capture dynamic patterns evolving throughout the record. The outcomes are promising, as DEAPS demonstrates a notable improvement of +10% over existing SOTA methods when just a few annotated records are presented to fit a Machine Learning (ML) model based on the learned representation.

Deep learning-based drug response prediction (DRP) methods can accelerate the drug discovery process and reduce research and development costs. Despite their high accuracy, generating regression-aware representations remains challenging for mainstream approaches. For instance, the representations are often disordered, aggregated, and overlapping, and they fail to characterize distinct samples effectively. This results in poor representation during the DRP task, diminishing generalizability and potentially leading to substantial costs during the drug discovery. In this paper, we propose CLDR, a contrastive learning framework with natural language supervision for the DRP. The CLDR converts regression labels into text, which is merged with the drug response caption as a second sample modality instead of the traditional modes, i.e., graphs and sequences. Simultaneously, a common-sense numerical knowledge graph is introduced to improve the continuous text representation. Our framework is validated using the genomics of drug sensitivity in cancer dataset with average performance increases ranging from 7.8% to 31.4%. Furthermore, experiments demonstrate that the proposed CLDR effectively maps samples with distinct label values into a high-dimensional space. In this space, the sample representations are scattered, significantly alleviating feature overlap. The code is available at: https://github.com/DrugD/CLDR.