搜索结果：backbone

Existing UDA pipelines fine-tune already well-trained backbone parameters for every new source-and-target pair, resulting in the number of training parameters and storage memory growing linearly with each new pair, and also preventing the reuse of these well-trained backbone parameters. Inspired by recent implications that existing backbones have textural biases, we propose making use of domain-specific textural bias for domain adaptation via visual reprogramming, namely VirDA. Instead of fine-tuning the full backbone, VirDA prepends a domain-specific visual reprogramming layer to the backbone. This layer produces visual prompts that act as an added textural bias to the input image, adapting its "style" to a target domain. To optimize these visual reprogramming layers, we use multiple objective functions that optimize the intra- and inter-domain distribution differences when domain-adapting visual prompts are applied. This process does not require modifying the backbone parameters, allowing the same backbone to be reused across different domains. We evaluate VirDA on Office-31 and obtain 92.8% mean accuracy with only 1.5M trainable parameters. VirDA surpasses PDA, the state-of-the-

Transfer learning has become an essential tool in modern computer vision, allowing practitioners to leverage backbones, pretrained on large datasets, to train successful models from limited annotated data. Choosing the right backbone is crucial, especially for small datasets, since final performance depends heavily on the quality of the initial feature representations. While prior work has conducted benchmarks across various datasets to identify universal top-performing backbones, we demonstrate that backbone effectiveness is highly dataset-dependent, especially in low-data scenarios where no single backbone consistently excels. To overcome this limitation, we introduce dataset-specific backbone selection as a new research direction and investigate its practical viability in low-data regimes. Since exhaustive evaluation is computationally impractical for large backbone pools, we formalize Vision Backbone Efficient Selection (VIBES) as the problem of searching for high-performing backbones under computational constraints. We define the solution space, propose several heuristics, and demonstrate VIBES feasibility for low-data image classification by performing experiments on four div

Recent advancements in LiDAR-based 3D object detection have significantly accelerated progress toward the realization of fully autonomous driving in real-world environments. Despite achieving high detection performance, most of the approaches still rely on a VGG-based or ResNet-based backbone for feature exploration, which increases the model complexity. Lightweight backbone design is well-explored for 2D object detection, but research on 3D object detection still remains limited. In this work, we introduce Dense Backbone, a lightweight backbone that combines the benefits of high processing speed, lightweight architecture, and robust detection accuracy. We adapt multiple SoTA 3d object detectors, such as PillarNet, with our backbone and show that with our backbone, these models retain most of their detection capability at a significantly reduced computational cost. To our knowledge, this is the first dense-layer-based backbone tailored specifically for 3D object detection from point cloud data. DensePillarNet, our adaptation of PillarNet, achieves a 29% reduction in model parameters and a 28% reduction in latency with just a 2% drop in detection accuracy on the nuScenes test set. F

Networks are essential for analyzing complex systems. However, their growing size necessitates backbone extraction techniques aimed at reducing their size while retaining critical features. In practice, selecting, implementing, and evaluating the most suitable backbone extraction method may be challenging. This paper introduces netbone, a Python package designed for assessing the performance of backbone extraction techniques in weighted networks. Its comparison framework is the standout feature of netbone. Indeed, the tool incorporates state-of-the-art backbone extraction techniques. Furthermore, it provides a comprehensive suite of evaluation metrics allowing users to evaluate different backbones techniques. We illustrate the flexibility and effectiveness of netbone through the US air transportation network analysis. We compare the performance of different backbone extraction techniques using the evaluation metrics. We also show how users can integrate a new backbone extraction method into the comparison framework. netbone is publicly available as an open-source tool, ensuring its accessibility to researchers and practitioners. Promoting standardized evaluation practices contribut

As networks grow in size and complexity, backbones become an essential network representation. Indeed, they provide a simplified yet informative overview of the underlying organization by retaining the most significant and structurally influential connections within a network. Network heterogeneity often results in complex and intricate structures, making it challenging to identify the backbone. In response, we introduce the Multilevel Backbone Extraction Framework, a novel approach that diverges from conventional backbone methodologies. This generic approach prioritizes the mesoscopic organization of networks. First, it splits the network into homogeneous-density components. Second, it extracts independent backbones for each component using any classical Backbone technique. Finally, the various backbones are combined. This strategy effectively addresses the heterogeneity observed in network groupings. Empirical investigations on real-world networks underscore the efficacy of the Multilevel Backbone approach in preserving essential network structures and properties. Experiments demonstrate its superiority over classical methods in handling network heterogeneity and enhancing networ

In practical scenarios, federated learning frequently necessitates training personalized models for each client using heterogeneous data. This paper proposes a backbone self-distillation approach to facilitate personalized federated learning. In this approach, each client trains its local model and only sends the backbone weights to the server. These weights are then aggregated to create a global backbone, which is returned to each client for updating. However, the client's local backbone lacks personalization because of the common representation. To solve this problem, each client further performs backbone self-distillation by using the global backbone as a teacher and transferring knowledge to update the local backbone. This process involves learning two components: the shared backbone for common representation and the private head for local personalization, which enables effective global knowledge transfer. Extensive experiments and comparisons with 12 state-of-the-art approaches demonstrate the effectiveness of our approach.

A backbone of a propositional CNF formula is a variable whose truth value is the same in every truth assignment that satisfies the formula. The notion of backbones for CNF formulas has been studied in various contexts. In this paper, we introduce local variants of backbones, and study the computational complexity of detecting them. In particular, we consider k-backbones, which are backbones for sub-formulas consisting of at most k clauses, and iterative k-backbones, which are backbones that result after repeated instantiations of k-backbones. We determine the parameterized complexity of deciding whether a variable is a k-backbone or an iterative k-backbone for various restricted formula classes, including Horn, definite Horn, and Krom. We also present some first empirical results regarding backbones for CNF-Satisfiability (SAT). The empirical results we obtain show that a large fraction of the backbones of structured SAT instances are local, in contrast to random instances, which appear to have few local backbones.

The metric backbone of a weighted graph is the union of all-pairs shortest paths. It is obtained by removing all edges $(u,v)$ that are not the shortest path between $u$ and $v$. In networks with well-separated communities, the metric backbone tends to preserve many inter-community edges, because these edges serve as bridges connecting two communities, but tends to delete many intra-community edges because the communities are dense. This suggests that the metric backbone would dilute or destroy the community structure of the network. However, this is not borne out by prior empirical work, which instead showed that the metric backbone of real networks preserves the community structure of the original network well. In this work, we analyze the metric backbone of a broad class of weighted random graphs with communities, and we formally prove the robustness of the community structure with respect to the deletion of all the edges that are not in the metric backbone. An empirical comparison of several graph sparsification techniques confirms our theoretical finding and shows that the metric backbone is an efficient sparsifier in the presence of communities.

Robust segmentation of clinically significant prostate cancer (csPCa) on multi-parametric MRI must tolerate frequent degradation of its most informative diffusion sequences. Multi-modal fusion commonly employs learned modality gating under the assumption that gates implement per-sample modality quality routing -- rarely tested directly. We ask how gating behaves across backbone architectures. We systematically analyze modality-isolated gated fusion (MIGF) for csPCa segmentation on two backbones (nnU-Net and Mamba) using PI-CAI (n=1500), with cross-cohort validation on Prostate158 (n=158): a factorial ablation over gating, modality dropout, and deep supervision under 5-fold cross-validation (180 trained models), plus a gate-weight and counterfactual analysis of 30 trained gating models. Modality gating is backbone-conditional. On nnU-Net, adding gating reduces the ranking score (marginal effect -0.037; gating configurations p<0.05), whereas on Mamba the gating-plus-dropout configuration improves it (+0.024, p=0.037). Gate-weight analysis explains this: nnU-Net gates collapse into a near-static modality prior (across-case SD 0.0033), while Mamba gates retain sample-dependent varia

Adaptations facilitate efficient training of large backbone models, including diffusion models for image generation and transformer-based language models. While various adaptation techniques enhance performance with minimal computational resources, limited adaptation data often leads to challenges in training. To address this, we focus on the enormous amount of backbone data used to pre-train the backbone models. We propose Backbone Augmented Training (BAT), a method that leverages backbone data to augment the adaptation dataset. First, we formulate and prove two mathematical key propositions: one establishes the validity of BAT, while the other identifies a condition under which BAT benefits adaptation. Furthermore, we introduce an advanced data selection scheme that satisfies these propositions and present ALBAT algorithm to implement this approach. ALBAT efficiently enhances adaptation training in both personalization and language generation tasks with scarce data.

In existing CNN based detectors, the backbone network is a very important component for basic feature extraction, and the performance of the detectors highly depends on it. In this paper, we aim to achieve better detection performance by building a more powerful backbone from existing backbones like ResNet and ResNeXt. Specifically, we propose a novel strategy for assembling multiple identical backbones by composite connections between the adjacent backbones, to form a more powerful backbone named Composite Backbone Network (CBNet). In this way, CBNet iteratively feeds the output features of the previous backbone, namely high-level features, as part of input features to the succeeding backbone, in a stage-by-stage fashion, and finally the feature maps of the last backbone (named Lead Backbone) are used for object detection. We show that CBNet can be very easily integrated into most state-of-the-art detectors and significantly improve their performances. For example, it boosts the mAP of FPN, Mask R-CNN and Cascade R-CNN on the COCO dataset by about 1.5 to 3.0 percent. Meanwhile, experimental results show that the instance segmentation results can also be improved. Specially, by sim

The $λ$-backbone coloring of the graph $G$ with backbone $H$ is a graph-coloring problem in which we are given a graph $G$ and a subgraph $H$, and we want to assign colors to vertices in such a way that the endpoints of every edge from $G$ have different colors, and the endpoints of every edge from $H$ are assigned colors which differ by at least $λ$. In this paper we pursue research on backbone coloring of bounded-degree graphs with well-known classes of backbones. Our result is an almost complete classification of problems in the form $BBC_λ(G, H) \le λ+ k$ for graphs with maximum degree $4$ and backbones from the following classes: paths, trees, matchings, and galaxies.

To realize a global quantum Internet, there is a need for communication between quantum subnetworks. To accomplish this task, there have been multiple design proposals for a quantum backbone network and quantum subnetworks. In this work, we elaborate on the design that uses entanglement and quantum teleportation to build the quantum backbone between packetized quantum networks. We design a network interface to interconnect packetized quantum networks with entanglement-based quantum backbone networks and, moreover, design a scheme to accomplish data transmission over this hybrid quantum network model. We analyze the use of various implementations of the backbone network, focusing our study on backbone networks that use satellite links to continuously distribute entanglement resources. For feasibility, we analyze various system parameters via simulation to benchmark the performance of the overall network.

Backbone members are recognized as essential parts of an organization, yet their role and mechanisms of functioning in networks are not fully understood. In this paper, we propose a new framework called Twotier to analyze the evolution of community sports organizations (CSOs) and the role of backbone members. Tier-one establishes a dynamic user interaction network based on grouping relationships, and weighted k-shell decomposition is used to select backbone members. We perform community detection and capture the evolution of two separate sub-networks: one formed by backbone members and the other formed by other members. In Tier-two, the sub-networks are abstracted, revealing a core-periphery structure in the organization where backbone members serve as bridges connecting all parts of the network. Our findings suggest that relying on backbone members can keep newcomers actively involved in rewarding activities, while non-rewarding activities solidify relations between backbone members.

Model compression techniques reduce the computational load and memory consumption of deep neural networks. After the compression operation, e.g. parameter pruning, the model is normally fine-tuned on the original training dataset to recover from the performance drop caused by compression. However, the training data is not always available due to privacy issues or other factors. In this work, we present a data-free fine-tuning approach for pruning the backbone of deep neural networks. In particular, the pruned network backbone is trained with synthetically generated images, and our proposed intermediate supervision to mimic the unpruned backbone's output feature map. Afterwards, the pruned backbone can be combined with the original network head to make predictions. We generate synthetic images by back-propagating gradients to noise images while relying on L1-pruning for the backbone pruning. In our experiments, we show that our approach is task-independent due to pruning only the backbone. By evaluating our approach on 2D human pose estimation, object detection, and image classification, we demonstrate promising performance compared to the unpruned model. Our code is available at ht

Object detection is a crucial component of autonomous driving, and many detection applications have been developed to address this task. These applications often rely on backbone architectures, which extract representation features from inputs to perform the object detection task. The quality of the features extracted by the backbone architecture can have a significant impact on the overall detection performance. Many researchers have focused on developing new and improved backbone architectures to enhance the efficiency and accuracy of object detection applications. While these backbone architectures have shown state-of-the-art performance on generic object detection datasets like MS-COCO and PASCAL-VOC, evaluating their performance under an autonomous driving environment has not been previously explored. To address this, our study evaluates three well-known autonomous vehicle datasets, namely KITTI, NuScenes, and BDD, to compare the performance of different backbone architectures on object detection tasks.

We report the recent derivation of the backbone exponent for 2D percolation. In contrast to previously known exactly solved percolation exponents, the backbone exponent is a transcendental number, which is a root of an elementary equation. We also report an exact formula for the probability that there are two disjoint paths of the same color crossing an annulus. The backbone exponent captures the leading asymptotic, while the other roots of the elementary equation capture the asymptotic of the remaining terms. This suggests that the backbone exponent is part of a conformal field theory (CFT) whose bulk spectrum contains this set of roots. Our approach is based on the coupling between SLE curves and Liouville quantum gravity (LQG), and the integrability of Liouville CFT that governs the LQG surfaces.

Contrastive Language-Image Pretraining (CLIP) stands out as a prominent method for image representation learning. Various neural architectures, spanning Transformer-based models like Vision Transformers (ViTs) to Convolutional Networks (ConvNets) like ResNets, are trained with CLIP and serve as universal backbones across diverse vision tasks. Despite utilizing the same data and training objectives, the effectiveness of representations learned by these architectures raises a critical question. Our investigation explores the differences in CLIP performance among these backbone architectures, revealing significant disparities in their classifications. Notably, normalizing these representations results in substantial performance variations. Our findings showcase a remarkable possible synergy between backbone predictions that could reach an improvement of over 20% through informed selection of the appropriate backbone. Moreover, we propose a simple, yet effective approach to combine predictions from multiple backbones, leading to a notable performance boost of up to 6.34\%. We will release the code for reproducing the results.

Tiny object detection has become an active area of research because images with tiny targets are common in several important real-world scenarios. However, existing tiny object detection methods use standard deep neural networks as their backbone architecture. We argue that such backbones are inappropriate for detecting tiny objects as they are designed for the classification of larger objects, and do not have the spatial resolution to identify small targets. Specifically, such backbones use max-pooling or a large stride at early stages in the architecture. This produces lower resolution feature-maps that can be efficiently processed by subsequent layers. However, such low-resolution feature-maps do not contain information that can reliably discriminate tiny objects. To solve this problem we design 'bottom-heavy' versions of backbones that allocate more resources to processing higher-resolution features without introducing any additional computational burden overall. We also investigate if pre-training these backbones on images of appropriate size, using CIFAR100 and ImageNet32, can further improve performance on tiny object detection. Results on TinyPerson and WiderFace show that