Similarity search, the task of identifying objects most similar to a given query object under a specific metric, has gathered significant attention due to its practical applications. However, the absence of coordinate information to accelerate similarity search and the high computational cost of measuring object similarity hinder the efficiency of existing CPU-based methods. Additionally, these methods struggle to meet the demand for high throughput data management. To address these challenges, we propose GTS, a GPU-based tree index designed for the parallel processing of similarity search in general metric spaces, where only the distance metric for measuring object similarity is known. The GTS index utilizes a pivot-based tree structure to efficiently prune objects and employs list tables to facilitate GPU computing. To efficiently manage concurrent similarity queries with limited GPU memory, we have developed a two-stage search method that combines batch processing and sequential strategies to optimize memory usage. The paper also introduces an effective update strategy for the proposed GPU-based index, encompassing streaming data updates and batch data updates. Additionally, we present a cost model to evaluate search performance. Extensive experiments on five real-life datasets demonstrate that GTS achieves efficiency gains of up to two orders of magnitude over existing CPU baselines and up to 20x efficiency improvements compared to state-of-the-art GPU-based methods.
相關內容
Unsupervised 3D object detection methods have emerged to leverage vast amounts of data efficiently without requiring manual labels for training. Recent approaches rely on dynamic objects for learning to detect objects but penalize the detections of static instances during training. Multiple rounds of (self) training are used in which detected static instances are added to the set of training targets; this procedure to improve performance is computationally expensive. To address this, we propose the method UNION. We use spatial clustering and self-supervised scene flow to obtain a set of static and dynamic object proposals from LiDAR. Subsequently, object proposals' visual appearances are encoded to distinguish static objects in the foreground and background by selecting static instances that are visually similar to dynamic objects. As a result, static and dynamic foreground objects are obtained together, and existing detectors can be trained with a single training. In addition, we extend 3D object discovery to detection by using object appearance-based cluster labels as pseudo-class labels for training object classification. We conduct extensive experiments on the nuScenes dataset and increase the state-of-the-art performance for unsupervised object discovery, i.e. UNION more than doubles the average precision to 33.9. The code will be made publicly available.
We present GSDeformer, a method that achieves free-form deformation on 3D Gaussian Splatting(3DGS) without requiring any architectural changes. Our method extends cage-based deformation, a traditional mesh deformation method, to 3DGS. This is done by converting 3DGS into a novel proxy point cloud representation, where its deformation can be used to infer the transformations to apply on the 3D gaussians making up 3DGS. We also propose an automatic cage construction algorithm for 3DGS to minimize manual work. Our method does not modify the underlying architecture of 3DGS. Therefore, any existing trained vanilla 3DGS can be easily edited by our method. We compare the deformation capability of our method against other existing methods, demonstrating the ease of use and comparable quality of our method, despite being more direct and thus easier to integrate with other concurrent developments on 3DGS.
Accurate, detailed, and high-frequent bathymetry, coupled with complex semantic content, is crucial for the undermapped shallow seabed areas facing intense climatological and anthropogenic pressures. Current methods exploiting remote sensing images to derive bathymetry or seabed classes mainly exploit non-open data. This lack of openly accessible benchmark archives prevents the wider use of deep learning methods in such applications. To address this issue, in this paper we present the MagicBathyNet, which is a benchmark dataset made up of image patches of Sentinel2, SPOT-6 and aerial imagery, bathymetry in raster format and annotations of seabed classes. MagicBathyNet is then exploited to benchmark state-of-the-art methods in learning-based bathymetry and pixel-based classification. Dataset, pre-trained weights, and code are publicly available at www.magicbathy.eu/magicbathynet.html.
Sparse models, including sparse Mixture-of-Experts (MoE) models, have emerged as an effective approach for scaling Transformer models. However, they often suffer from computational inefficiency since a significant number of parameters are unnecessarily involved in computations via multiplying values by zero or low activation values. To address this issue, we present \tool, a novel MoE designed to enhance both the efficacy and efficiency of sparse MoE models. \tool leverages small experts and a threshold-based router to enable tokens to selectively engage only essential parameters. Our extensive experiments on language modeling and machine translation tasks demonstrate that \tool can enhance model performance while decreasing the computation load at MoE layers by over 50\% without sacrificing performance. Furthermore, we present the versatility of \tool by applying it to dense models, enabling sparse computation during inference. We provide a comprehensive analysis and make our code available at //github.com/ysngki/XMoE.
Neural implicit representations have recently demonstrated considerable potential in the field of visual simultaneous localization and mapping (SLAM). This is due to their inherent advantages, including low storage overhead and representation continuity. However, these methods necessitate the size of the scene as input, which is impractical for unknown scenes. Consequently, we propose NeB-SLAM, a neural block-based scalable RGB-D SLAM for unknown scenes. Specifically, we first propose a divide-and-conquer mapping strategy that represents the entire unknown scene as a set of sub-maps. These sub-maps are a set of neural blocks of fixed size. Then, we introduce an adaptive map growth strategy to achieve adaptive allocation of neural blocks during camera tracking and gradually cover the whole unknown scene. Finally, extensive evaluations on various datasets demonstrate that our method is competitive in both mapping and tracking when targeting unknown environments.
Depth estimation based on stereo matching is a classic but popular computer vision problem, which has a wide range of real-world applications. Current stereo matching methods generally adopt the deep Siamese neural network architecture, and have achieved impressing performance by constructing feature matching cost volumes and using 3D convolutions for cost aggregation. However, most existing methods suffer from large number of parameters and slow running time due to the sequential use of 3D convolutions. In this paper, we propose Ghost-Stereo, a novel end-to-end stereo matching network. The feature extraction part of the network uses the GhostNet to form a U-shaped structure. The core of Ghost-Stereo is a GhostNet feature-based cost volume enhancement (Ghost-CVE) module and a GhostNet-inspired lightweight cost volume aggregation (Ghost-CVA) module. For the Ghost-CVE part, cost volumes are constructed and fused by the GhostNet-based features to enhance the spatial context awareness. For the Ghost-CVA part, a lightweight 3D convolution bottleneck block based on the GhostNet is proposed to reduce the computational complexity in this module. By combining with the context and geometry fusion module, a classical hourglass-shaped cost volume aggregate structure is constructed. Ghost-Stereo achieves a comparable performance than state-of-the-art real-time methods on several publicly benchmarks, and shows a better generalization ability.
Autonomously exploring the unknown physical properties of novel objects such as stiffness, mass, center of mass, friction coefficient, and shape is crucial for autonomous robotic systems operating continuously in unstructured environments. We introduce a novel visuo-tactile based predictive cross-modal perception framework where initial visual observations (shape) aid in obtaining an initial prior over the object properties (mass). The initial prior improves the efficiency of the object property estimation, which is autonomously inferred via interactive non-prehensile pushing and using a dual filtering approach. The inferred properties are then used to enhance the predictive capability of the cross-modal function efficiently by using a human-inspired `surprise' formulation. We evaluated our proposed framework in the real-robotic scenario, demonstrating superior performance.
Self-supervised contrastive learning, which directly extracts inherent data correlations from unlabeled data, has been widely utilized to mitigate the data sparsity issue in sequential recommendation. The majority of existing methods create different augmented views of the same user sequence via random augmentation, and subsequently minimize their distance in the embedding space to enhance the quality of user representations. However, random augmentation often disrupts the semantic information and interest evolution pattern inherent in the user sequence, leading to the generation of semantically distinct augmented views. Promoting similarity of these semantically diverse augmented sequences can render the learned user representations insensitive to variations in user preferences and interest evolution, contradicting the core learning objectives of sequential recommendation. To address this issue, we leverage the inherent characteristics of sequential recommendation and propose the use of context information to generate more reasonable augmented positive samples. Specifically, we introduce a context-aware diffusion-based contrastive learning method for sequential recommendation. Given a user sequence, our method selects certain positions and employs a context-aware diffusion model to generate alternative items for these positions with the guidance of context information. These generated items then replace the corresponding original items, creating a semantically consistent augmented view of the original sequence. Additionally, to maintain representation cohesion, item embeddings are shared between the diffusion model and the recommendation model, and the entire framework is trained in an end-to-end manner. Extensive experiments on five benchmark datasets demonstrate the superiority of our proposed method.
Most object recognition approaches predominantly focus on learning discriminative visual patterns while overlooking the holistic object structure. Though important, structure modeling usually requires significant manual annotations and therefore is labor-intensive. In this paper, we propose to "look into object" (explicitly yet intrinsically model the object structure) through incorporating self-supervisions into the traditional framework. We show the recognition backbone can be substantially enhanced for more robust representation learning, without any cost of extra annotation and inference speed. Specifically, we first propose an object-extent learning module for localizing the object according to the visual patterns shared among the instances in the same category. We then design a spatial context learning module for modeling the internal structures of the object, through predicting the relative positions within the extent. These two modules can be easily plugged into any backbone networks during training and detached at inference time. Extensive experiments show that our look-into-object approach (LIO) achieves large performance gain on a number of benchmarks, including generic object recognition (ImageNet) and fine-grained object recognition tasks (CUB, Cars, Aircraft). We also show that this learning paradigm is highly generalizable to other tasks such as object detection and segmentation (MS COCO). Project page: //github.com/JDAI-CV/LIO.
Recently, ensemble has been applied to deep metric learning to yield state-of-the-art results. Deep metric learning aims to learn deep neural networks for feature embeddings, distances of which satisfy given constraint. In deep metric learning, ensemble takes average of distances learned by multiple learners. As one important aspect of ensemble, the learners should be diverse in their feature embeddings. To this end, we propose an attention-based ensemble, which uses multiple attention masks, so that each learner can attend to different parts of the object. We also propose a divergence loss, which encourages diversity among the learners. The proposed method is applied to the standard benchmarks of deep metric learning and experimental results show that it outperforms the state-of-the-art methods by a significant margin on image retrieval tasks.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191