AAAI 2025 论文列表

This paper explores the application of computer vision technology as a proactive solution to prevent Road Traffic Accidents in Nigeria. By leveraging machine learning algorithms and real-time video analysis, computer vision can reduce incidents caused by human error. The research focuses on designing an autonomous but elaborate system that monitors traffic patterns, road irregularities and triggers automated interventions when risky conditions are detected. The aim is to suggest that computer vision can be pivotal in enhancing road safety and reducing traffic-related fatalities in Nigeria.

3D Gaussian Splatting (3DGS) has shown promising performance in novel view synthesis. Previous methods adapt it to obtaining surfaces of either individual 3D objects or within limited scenes. In this paper, we make the first attempt to tackle the challenging task of large-scale scene surface reconstruction. This task is particularly difficult due to the high GPU memory consumption, different levels of details for geometric representation, and noticeable inconsistencies in appearance. To this end, we propose GigaGS, the first work for high-quality surface reconstruction for large-scale scenes using 3DGS. GigaGS first applies a partitioning strategy based on the mutual visibility of spatial regions, which effectively grouping cameras for parallel processing. To enhance the quality of the surface, we also propose novel multi-view photometric and geometric consistency constraints based on Level-of-Detail representation. In doing so, our method can reconstruct detailed surface structures. Comprehensive experiments are conducted on various datasets. The consistent improvement demonstrates the superiority of GigaGS.

Information diffusion prediction (IDP) is a pivotal task for understanding the dynamics of information propagation within social networks. Conventional models typically adhere to a fixed learning-based paradigm, where the trained prediction model remains static during the inference phase. This paradigm presupposes that the data is independent and identically distributed, an assumption that may not hold true due to the inherently open nature of social media and the uncertainty and variability in user behavior. In this paper, we address the novel problem of out-of-distribution (OOD) shifts within IDP tasks and propose a new test-time training-based model for multi-scale IDP tasks, named Ghidorah. Our approach focuses on adapting a subset of model parameters to accommodate the unique characteristics of test samples through self-supervised learning (SSL) tasks. Ghidorah comprises three components: the macroscopic prediction branch, the microscopic prediction branch, and the auxiliary SSL branch. The auxiliary SSL task employs a masked autoencoder-based loss to fine-tune the model for specific test samples prior to prediction. Furthermore, Ghidorah integrates invariant learning to capture robust representations while mitigating spurious correlations. To our knowledge, Ghidorah is the first work to introduce a test-time training framework specifically designed to address the critical yet often overlooked OOD challenges in IDP. Experimental results across several benchmark datasets validate the superiority of our approach.

Inconsistency is a common problem in knowledge, and so there is a need to analyse it. Inconsistency measures assess its severity, but there is a more basic question: "where is the inconsistency?". Typically, not all subsets of a knowledgebase are causing the inconsistency, and minimal inconsistent sets have been the standard way to localise the germane ones, even though there are shortcomings in some scenarios. Recently, ⋆-conflicts were proposed as a more suitable definition to localise inconsistency when considering a method to repair it. But in general there is no way to tell what is a sensible definition to capture the germane conflicts. This work provides a set of desirable properties to assess definitions for germane conflicts. Also, a new conflict definition, based on substitution, is presented and evaluated via the proposed properties, and the related computational complexity is analysed.

Point cloud salient object detection has attracted the attention of researchers in recent years. Since existing works do not fully utilize the geometry context of 3D objects, blurry boundaries are generated when segmenting objects with complex backgrounds. In this paper, we propose a geometry-aware 3D salient object detection network that explicitly clusters points into superpoints to enhance the geometric boundaries of objects, thereby segmenting complete objects with clear boundaries. Specifically, we first propose a simple yet effective superpoint partition module to cluster points into superpoints. In order to improve the quality of superpoints, we present a point cloud class-agnostic loss to learn discriminative point features for clustering superpoints from the object. After obtaining superpoints, we then propose a geometry enhancement module that utilizes superpoint-point attention to aggregate geometric information into point features for predicting the salient map of the object with clear boundaries. Extensive experiments show that our method achieves new state-of-the-art performance on the PCSOD dataset.

In the geospatial domain, universal representation models are significantly less prevalent than their extensive use in natural language processing and computer vision. This discrepancy arises primarily from the high costs associated with the input of existing representation models, which often require street views and mobility data. To address this, we develop a novel, training-free method that leverages large language models (LLMs) and auxiliary map data from OpenStreetMap to derive geolocation representations (LLMGeovec). LLMGeovec can represent the geographic semantics of city, country, and global scales, which acts as a generic enhancer for spatio-temporal learning. Specifically, by direct feature concatenation, we introduce a simple yet effective paradigm for enhancing multiple spatio-temporal tasks including geographic prediction (GP), long-term time series forecasting (LTSF), and graph-based spatio-temporal forecasting (GSTF). LLMGeovec can seamlessly integrate into a wide spectrum of spatio-temporal learning models, providing immediate enhancements. Experimental results demonstrate that LLMGeovec achieves global coverage and significantly boosts the performance of leading GP, LTSF, and GSTF models.

Tabular data poses unique challenges due to its heterogeneous nature, combining both continuous and categorical variables. Existing approaches often struggle to effectively capture the underlying structure and relationships within such data. We propose GFTab (Geodesic Flow Kernels for Semi-Supervised Learning on Mixed-Variable Tabular Dataset), a semi-supervised framework specifically designed for tabular datasets. GFTab incorporates three key innovations: 1) Variable-specific corruption methods tailored to the distinct properties of continuous and categorical variables, 2) A Geodesic flow kernel based similarity measure to capture geometric changes between corrupted inputs, and 3) Tree-based embedding to leverage hierarchical relationships from available labeled data. To rigorously evaluate GFTab, we curate a comprehensive set of 21 tabular datasets spanning various domains, sizes, and variable compositions. Our experimental results show that GFTab outperforms existing ML/DL models across many of these datasets, particularly in settings with limited labeled data.

The problem of forecasting spatiotemporal events such as crimes and accidents is crucial to public safety and city management. Besides accuracy, interpretability is also a key requirement for spatiotemporal forecasting models to justify the decisions. Merely presenting predicted scores fails to convince the public and does not contribute to future urban planning. Interpretation of the spatiotemporal forecasting mechanism is, however, challenging due to the complexity of multi-source spatiotemporal features, the non-intuitive nature of spatiotemporal patterns for non-expert users, and the presence of spatial heterogeneity in the data. Currently, no existing deep learning model intrinsically interprets the complex predictive process learned from multi-source spatiotemporal features. To bridge the gap, we propose GeoPro-Net, an intrinsically interpretable spatiotemporal model for spatiotemporal event forecasting problems. GeoPro-Net introduces a novel Geo-concept convolution operation, which employs statistical tests to extract predictive patterns in the input as "Geo-concepts'', and condenses the "Geo-concept-encoded'' input through interpretable channel fusion and geographic-based pooling. In addition, GeoPro-Net learns different sets of prototypes of concepts inherently, and projects them to real-world cases for interpretation. Comprehensive experiments and case studies on four real-world datasets demonstrate that GeoPro-Net provides better interpretability while still achieving competitive prediction performance compared with state-of-the-art baselines.

Point-of-Interest (POI) recommendation plays an important role in a wide range of location-based social network ap- plications, aiming to accurately predicting users’ next visits based on their historical check-in records. Previous efforts have primarily focused on the modifications of existing sequential models, neglecting the fact that POI visiting sequences typically involve continuous state transformation of geographical and intention signals. Additionally, the diverse time span between check-ins require the model to prop- erly recognize user’s multi-granular preference. While recent advances of State Space Model (SSM) have revealed their potential in handling intricate temporal signals, we propose a state-based model that is tailored for spatio-temporal POI sequences. On top of traditional SSMs that are typically limited to linear sequences like Mamba, we propose GeoMamba, which customizes the model states to accommodate the spatio-temporal sequences, especially fitting for POI recommendations. Specifically, while the approximation operator HiPPO sets the foundation of linear SSMs, we introduce a novel GaPPO operator that extends the model’s state space into graph-represented geographical domains. This innovation allows us to construct locational SSM encoders that seamlessly integrate users’ spatio-temporal characteristics. The sequence-aware outputs of GeoMamba are further processed to generate multi-scale behavior representations. Extensive experimental results illustrate the superiority of GeoMamba over several state-of-the-art baselines.

Bird's-Eye-View (BEV) representation has emerged as a mainstream paradigm for multi-view 3D object detection, demonstrating impressive perceptual capabilities. However, existing methods overlook the geometric quality of BEV representation, leaving it in a low-resolution state and failing to restore the authentic geometric information of the scene. In this paper, we identify the drawbacks of previous approaches that limit the geometric quality of BEV representation and propose Radial-Cartesian BEV Sampling (RC-Sampling), which outperforms other feature transformation methods in efficiently generating high-resolution dense BEV representation to restore fine-grained geometric information. Additionally, we design a novel In-Box Label to substitute the traditional depth label generated from the LiDAR points. This label reflects the actual geometric structure of objects rather than just their surfaces, injecting real-world geometric information into the BEV representation. In conjunction with the In-Box Label, Centroid-Aware Inner Loss (CAI Loss) is developed to capture the inner geometric structure of objects. Finally, we integrate the aforementioned modules into a novel multi-view 3D object detector, dubbed GeoBEV, which achieves a state-of-the-art result of 66.2% NDS on the nuScenes test set.

Modeling geospatial tabular data with deep learning has become a promising alternative to traditional statistical and machine learning approaches. However, existing deep learning models often face challenges related to scalability and flexibility as datasets grow. To this end, this paper introduces GeoAggregator, an efficient and lightweight algorithm based on transformer architecture designed specifically for geospatial tabular data modeling. GeoAggregators explicitly account for spatial autocorrelation and spatial heterogeneity through Gaussian-biased local attention and global positional awareness. Additionally, we introduce a new attention mechanism that uses the Cartesian product to manage the size of the model while maintaining strong expressive power. We benchmark GeoAggregator against spatial statistical models, XGBoost, and several state-of-the-art geospatial deep learning methods using both synthetic and empirical geospatial datasets. The results demonstrate that GeoAggregators achieve the best or second-best performance compared to their competitors on nearly all datasets. GeoAggregator's efficiency is underscored by its reduced model size, making it both scalable and lightweight. Moreover, ablation experiments offer insights into the effectiveness of the Gaussian bias and Cartesian attention mechanism, providing recommendations for further optimizing the GeoAggregator's performance.

Learning-based compression shows competitive compression ratios for genomics data. It often includes three types of compressors: static, adaptive and semi-adaptive. However, these existing compressors suffer from inferior compression ratios or throughput, and adaptive compressors also faces model cold-start problems. To address these issues, we propose DeepGeCo, a novel genomics data lossless adaptive compression framework with (s,k)-mer encoding and deep neural networks, involving three compression modes (MINI for static, PLUS for adaptive, ULTRA for semi-adaptive) for flexible requirements of compression ratios or throughput. In DeepGeCo, (1) we develop BiGRU and Transformer as the backbone to build Warm-Start and Supporter models in terms of cold-start problems. (2) We introduce (s,k)-mer encoding to pre-process genomics data before feeding it into the DNN model for improve model throughput, and we propose a new metric - Ranking of Throughput and Compression Ratio (RTCR) for effective encoding parameters selection. (3) We design a threshold controller and a probabilistic mixer within the backbone to balance compression ratios and model throughput. Experiments on 10 real-world datasets show that DeepGeCo's three compression modes improve up to a 22.949X average throughput and up to a 31.095% average compression ratio improvement while occupying low CPU or GPU memory.

Large-scale text-guided image diffusion models have demonstrated remarkable results in text-to-image (T2I) generation. However, applying these models to synthesize textures for 3D geometries remains challenging due to the domain gap between 2D images and textures on a 3D surface. Early works that used a projecting-inpainting approach managed to preserve generation diversity, but often resulted in noticeable artifacts and style inconsistencies. While recent methods have attempted to address these inconsistencies, they often introduce other issues, such as blurring, over-saturation, or over-smoothing. To overcome these challenges, we propose a novel text-to-texture synthesis framework that takes advantage of pre-trained diffusion models. We introduce a local attention reweighing mechanism in the self-attention layers to guide the model in focusing on spatial-correlated patches across different views, thereby enhancing local details while preserving cross-view consistency. Additionally, we propose a novel latent space merge pipeline, which further ensures consistency across different viewpoints without sacrificing too much diversity. Our method significantly outperforms existing state-of-the-art techniques in terms of texture consistency and visual quality, while delivering results much faster than distillation-based methods. Importantly, our framework does not require additional training or fine-tuning, making it highly adaptable to a wide range of models available on public platforms.

Video moment retrieval (VMR) aims to locate the most likely video moment(s) corresponding to a text query in untrimmed videos. Training of existing methods is limited by the lack of diverse and generalisable VMR datasets, hindering their ability to generalise moment-text associations to queries containing novel semantic concepts (unseen both visually and textually in a training source domain). For model generalisation to novel semantics, existing methods rely heavily on assuming to have access to both video and text sentence pairs from a target domain in addition to the source domain pair-wise training data. This is neither practical nor scalable. In this work, we introduce a more generalisable approach by assuming only text sentences describing new semantics are available in model training without having seen any videos from a target domain. To that end, we propose a Fine-grained Video Editing framework, termed FVE, that explores generative video diffusion to facilitate fine-grained video editing from the seen source concepts to the unseen target sentences consisting of new concepts. This enables generative hypotheses of unseen video moments corresponding to the novel concepts in the target domain. This fine-grained generative video diffusion retains the original video structure and subject specifics from the source domain while introducing semantic distinctions of unseen novel vocabularies in the target domain. A critical challenge is how to enable this generative fine-grained diffusion process to be meaningful in optimising VMR, more than just synthesising visually pleasing videos. We solve this problem by introducing a hybrid selection mechanism that integrates three quantitative metrics to selectively incorporate synthetic video moments (novel video hypotheses) as enlarged additions to the original source training data, whilst minimising potential detrimental noise or unnecessary repetitions in the novel synthetic videos harmful to VMR learning. Experiments on three datasets demonstrate the effectiveness of FVE to unseen novel semantic video moment retrieval tasks

Autonomous driving is a challenging task that requires perceiving and understanding the surrounding environment for safe trajectory planning. While existing vision-based end-to-end models have achieved promising results, these methods are still facing the challenges of vision understanding, decision reasoning and scene generalization. To solve these issues, a generative planning with 3D-vision language pre-training model named GPVL is proposed for end-to-end autonomous driving. The proposed paradigm has two significant aspects. On one hand, a 3D-vision language pre-training module is designed to bridge the gap between visual perception and linguistic understanding in the bird's eye view. On the other hand, a cross-modal language model is introduced to generate reasonable planning with perception and navigation information in an auto-regressive manner. Experiments on the challenging nuScenes dataset demonstrate that the proposed scheme achieves excellent performances compared with state-of-the-art methods. Besides, the proposed GPVL presents strong generalization ability and real-time potential when handling high-level commands in various scenarios. It is believed that the effective, robust and efficient performance of GPVL is crucial for the practical application of future autonomous driving systems.

Rapid advancements in medical image segmentation performance have been significantly driven by the development of Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs). These models follow discriminative pixel-wise classification learning paradigm and often have limited ability to generalize across diverse medical imaging datasets. In this manuscript, we introduce Generative Medical Segmentation (GMS), a novel generative approach to perform image segmentation. GMS employs a robust pre-trained vision foundation model to extract latent representations for images and corresponding ground truth masks, followed by a lightweight model that learns a mapping function from the image to the mask in the latent space. Once trained, the model can generate estimated segmentation masks using the pre-trained vision foundation model to decode the predicted latent mask representation back into image space. The design of GMS leads to fewer trainable parameters in the model, reducing the risk of overfitting and enhancing its generalization capability. Our experimental analysis across five open-source datasets in different medical imaging domains demonstrates GMS outperforms existing discriminative and generative segmentation models. Furthermore, GMS is able to generalize well across datasets of the same imaging modality from different centers. Our experiments suggest GMS offers a scalable and effective solution for medical image segmentation.

This paper investigates the application of Generative Flow Networks (GFlowNets) to lead optimization in drug discovery. GFlowNets provide a novel framework for generating diverse molecular structures while optimizing for desired properties, addressing the limitations of traditional methods in exploring vast chemical spaces. We adapt GFlowNets to incrementally modify lead compounds, integrating domain-specific heuristics to guide the generation process. Our method employs the trajectory balance objective on a graph neural network (GNN), to learn a policy that samples fragments based on a multi-objective reward. The reward function ensures increase in cell permeability and similarity to the starting molecule. The results on benchmark datasets of activity cliffs demonstrate that GFlowNets can generate diverse modifications, producing optimized candidate molecules with improvement in cell permeability. This work can be extended with other pharmacokinetic properties for lead optimization in early-stage drug development, potentially accelerating the discovery of novel therapeutics.

Motion planning is a crucial component in autonomous driving. State-of-the-art motion planners are trained on meticulously curated datasets, which are not only expensive to annotate but also insufficient in capturing rarely seen critical scenarios. Failing to account for such scenarios poses a significant risk to motion planners and may lead to incidents during testing. An intuitive solution is to manually compose such scenarios by programming and executing a simulator (e.g., CARLA). However, this approach incurs substantial human costs. Motivated by this, we propose an inexpensive method for generating diverse critical traffic scenarios to train more robust motion planners. First, we represent traffic scenarios as scripts, which are then used by the simulator to generate traffic scenarios. Next, we develop a method that accepts user-specified text descriptions, which a Large Language Model translates into scripts using in-context learning. The output scripts are sent to the simulator that produces the corresponding traffic scenarios. As our method can generate abundant safety-critical traffic scenarios, we use them as synthetic training data for motion planners. To demonstrate the value of generated scenarios, we train existing motion planners on our synthetic data, real-world datasets, and a combination of both. Our experiments show that motion planners trained with our data significantly outperform those trained solely on real-world data, showing the usefulness of our synthetic data and the effectiveness of our data generation method.

Unsupervised federated learning for cross-modal retrieval has received increasing attention in recent years as it can free the requirement for annotations and avoid uploading original clients’ data to servers. Most existing methods focus on how to learn better local models and their aggregation to overcome data distribution drift across clients. Unlike prior works, we propose to address the data distribution problem by generating synthetic data, which can benefit existing federated learning methods. Specifically, we train a WGAN generator with three newly designed loss constraints on each client to improve the quality of the generated data. We first compute cluster prototypes to address the problem of lack of labels. Then, a direct contrastive loss between generated image and text features, an indirect contrastive loss with reference to cluster prototypes, and a Jensen-Shannon Divergence (JSD) loss also with reference to cluster prototypes work together to constrain the WGAN. The locally trained generators and local prototypes are sent to the server to generate and filter synthetic data with consideration of data distribution across all clients. The filtered data are used to train the aggregated global retrieval model, which is later sent to clients. The final global model becomes robust to all clients after several rounds of client-server iteration. Extensive experiments using four baselines across three datasets demonstrate that our method performs favourably against state-of-the-art methods.