PyTorchKR Papers

최신 논문

총 3,796편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2210.02843공식 코드 · Papers with Code

CIR-Net: Cross-modality Interaction and Refinement for RGB-D Salient Object Detection

Focusing on the issue of how to effectively capture and utilize cross-modality information in RGB-D salient object detection (SOD) task, we present a convolutional neural network (CNN) model, named CIR-Net, based on the novel cross-modality interaction and refinement. For the cross-modality interaction, 1) a progressive attention guided integration unit is proposed to sufficiently integrate RGB-D feature representations in the encoder stage, and 2) a convergence aggregation structure is proposed, which flows the RGB and depth decoding features into the corresponding RGB-D decoding streams via an importance gated fusion unit in the decoder stage. For the cross-modality refinement, we insert a refinement middleware structure between the encoder and the decoder, in which the RGB, depth, and RGB-D encoder features are further refined by successively using a self-modality attention refinement unit and a cross-modality weighting refinement unit. At last, with the gradually refined features, we predict the saliency map in the decoder stage. Extensive experiments on six popular RGB-D SOD benchmarks demonstrate that our network outperforms the state-of-the-art saliency detectors both qualitatively and quantitatively.

arXiv:2210.02808공식 코드 · Papers with Code

Effective Self-supervised Pre-training on Low-compute Networks without Distillation

Despite the impressive progress of self-supervised learning (SSL), its applicability to low-compute networks has received limited attention. Reported performance has trailed behind standard supervised pre-training by a large margin, barring self-supervised learning from making an impact on models that are deployed on device. Most prior works attribute this poor performance to the capacity bottleneck of the low-compute networks and opt to bypass the problem through the use of knowledge distillation (KD). In this work, we revisit SSL for efficient neural networks, taking a closer at what are the detrimental factors causing the practical limitations, and whether they are intrinsic to the self-supervised low-compute setting. We find that, contrary to accepted knowledge, there is no intrinsic architectural bottleneck, we diagnose that the performance bottleneck is related to the model complexity vs regularization strength trade-off. In particular, we start by empirically observing that the use of local views can have a dramatic impact on the effectiveness of the SSL methods. This hints at view sampling being one of the performance bottlenecks for SSL on low-capacity networks. We hypothesize that the view sampling strategy for large neural networks, which requires matching views in very diverse spatial scales and contexts, is too demanding for low-capacity architectures. We systematize the design of the view sampling mechanism, leading to a new training methodology that consistently improves the performance across different SSL methods (e.g. MoCo-v2, SwAV, DINO), different low-size networks (e.g. MobileNetV2, ResNet18, ResNet34, ViT-Ti), and different tasks (linear probe, object detection, instance segmentation and semi-supervised learning). Our best models establish a new state-of-the-art for SSL methods on low-compute networks despite not using a KD loss term.

arXiv:2210.02077공식 코드 · Papers with Code

Exploring The Role of Mean Teachers in Self-supervised Masked Auto-Encoders

Masked image modeling (MIM) has become a popular strategy for self-supervised learning~(SSL) of visual representations with Vision Transformers. A representative MIM model, the masked auto-encoder (MAE), randomly masks a subset of image patches and reconstructs the masked patches given the unmasked patches. Concurrently, many recent works in self-supervised learning utilize the student/teacher paradigm which provides the student with an additional target based on the output of a teacher composed of an exponential moving average (EMA) of previous students. Although common, relatively little is known about the dynamics of the interaction between the student and teacher. Through analysis on a simple linear model, we find that the teacher conditionally removes previous gradient directions based on feature similarities which effectively acts as a conditional momentum regularizer. From this analysis, we present a simple SSL method, the Reconstruction-Consistent Masked Auto-Encoder (RC-MAE) by adding an EMA teacher to MAE. We find that RC-MAE converges faster and requires less memory usage than state-of-the-art self-distillation methods during pre-training, which may provide a way to enhance the practicality of prohibitively expensive self-supervised learning of Vision Transformer models. Additionally, we show that RC-MAE achieves more robustness and better performance compared to MAE on downstream tasks such as ImageNet-1K classification, object detection, and instance segmentation.

arXiv:2210.02021공식 코드 · Papers with Code

Exploring Effective Knowledge Transfer for Few-shot Object Detection

Recently, few-shot object detection~(FSOD) has received much attention from the community, and many methods are proposed to address this problem from a knowledge transfer perspective. Though promising results have been achieved, these methods fail to achieve shot-stable:~methods that excel in low-shot regimes are likely to struggle in high-shot regimes, and vice versa. We believe this is because the primary challenge of FSOD changes when the number of shots varies. In the low-shot regime, the primary challenge is the lack of inner-class variation. In the high-shot regime, as the variance approaches the real one, the main hindrance to the performance comes from misalignment between learned and true distributions. However, these two distinct issues remain unsolved in most existing FSOD methods. In this paper, we propose to overcome these challenges by exploiting rich knowledge the model has learned and effectively transferring them to the novel classes. For the low-shot regime, we propose a distribution calibration method to deal with the lack of inner-class variation problem. Meanwhile, a shift compensation method is proposed to compensate for possible distribution shift during fine-tuning. For the high-shot regime, we propose to use the knowledge learned from ImageNet as guidance for the feature learning in the fine-tuning stage, which will implicitly align the distributions of the novel classes. Although targeted toward different regimes, these two strategies can work together to further improve the FSOD performance. Experiments on both the VOC and COCO benchmarks show that our proposed method can significantly outperform the baseline method and produce competitive results in both low-shot settings (shot<5) and high-shot settings (shot>=5). Code is available at https://github.com/JulioZhao97/EffTrans_Fsdet.git.

arXiv:2210.01820공식 코드 · Papers with Code

MOAT: Alternating Mobile Convolution and Attention Brings Strong Vision Models

This paper presents MOAT, a family of neural networks that build on top of MObile convolution (i.e., inverted residual blocks) and ATtention. Unlike the current works that stack separate mobile convolution and transformer blocks, we effectively merge them into a MOAT block. Starting with a standard Transformer block, we replace its multi-layer perceptron with a mobile convolution block, and further reorder it before the self-attention operation. The mobile convolution block not only enhances the network representation capacity, but also produces better downsampled features. Our conceptually simple MOAT networks are surprisingly effective, achieving 89.1% / 81.5% top-1 accuracy on ImageNet-1K / ImageNet-1K-V2 with ImageNet22K pretraining. Additionally, MOAT can be seamlessly applied to downstream tasks that require large resolution inputs by simply converting the global attention to window attention. Thanks to the mobile convolution that effectively exchanges local information between pixels (and thus cross-windows), MOAT does not need the extra window-shifting mechanism. As a result, on COCO object detection, MOAT achieves 59.2% box AP with 227M model parameters (single-scale inference, and hard NMS), and on ADE20K semantic segmentation, MOAT attains 57.6% mIoU with 496M model parameters (single-scale inference). Finally, the tiny-MOAT family, obtained by simply reducing the channel sizes, also surprisingly outperforms several mobile-specific transformer-based models on ImageNet. The tiny-MOAT family is also benchmarked on downstream tasks, serving as a baseline for the community. We hope our simple yet effective MOAT will inspire more seamless integration of convolution and self-attention. Code is publicly available.

arXiv:2210.02443공식 코드 · Papers with Code

Time Will Tell: New Outlooks and A Baseline for Temporal Multi-View 3D Object Detection

While recent camera-only 3D detection methods leverage multiple timesteps, the limited history they use significantly hampers the extent to which temporal fusion can improve object perception. Observing that existing works' fusion of multi-frame images are instances of temporal stereo matching, we find that performance is hindered by the interplay between 1) the low granularity of matching resolution and 2) the sub-optimal multi-view setup produced by limited history usage. Our theoretical and empirical analysis demonstrates that the optimal temporal difference between views varies significantly for different pixels and depths, making it necessary to fuse many timesteps over long-term history. Building on our investigation, we propose to generate a cost volume from a long history of image observations, compensating for the coarse but efficient matching resolution with a more optimal multi-view matching setup. Further, we augment the per-frame monocular depth predictions used for long-term, coarse matching with short-term, fine-grained matching and find that long and short term temporal fusion are highly complementary. While maintaining high efficiency, our framework sets new state-of-the-art on nuScenes, achieving first place on the test set and outperforming previous best art by 5.2% mAP and 3.7% NDS on the validation set. Code will be released $\href{https://github.com/Divadi/SOLOFusion}{here.}$

arXiv:2210.02093공식 코드 · Papers with Code

Centralized Feature Pyramid for Object Detection

Visual feature pyramid has shown its superiority in both effectiveness and efficiency in a wide range of applications. However, the existing methods exorbitantly concentrate on the inter-layer feature interactions but ignore the intra-layer feature regulations, which are empirically proved beneficial. Although some methods try to learn a compact intra-layer feature representation with the help of the attention mechanism or the vision transformer, they ignore the neglected corner regions that are important for dense prediction tasks. To address this problem, in this paper, we propose a Centralized Feature Pyramid (CFP) for object detection, which is based on a globally explicit centralized feature regulation. Specifically, we first propose a spatial explicit visual center scheme, where a lightweight MLP is used to capture the globally long-range dependencies and a parallel learnable visual center mechanism is used to capture the local corner regions of the input images. Based on this, we then propose a globally centralized regulation for the commonly-used feature pyramid in a top-down fashion, where the explicit visual center information obtained from the deepest intra-layer feature is used to regulate frontal shallow features. Compared to the existing feature pyramids, CFP not only has the ability to capture the global long-range dependencies, but also efficiently obtain an all-round yet discriminative feature representation. Experimental results on the challenging MS-COCO validate that our proposed CFP can achieve the consistent performance gains on the state-of-the-art YOLOv5 and YOLOX object detection baselines.

arXiv:2210.02088공식 코드 · Papers with Code

WUDA: Unsupervised Domain Adaptation Based on Weak Source Domain Labels

Unsupervised domain adaptation (UDA) for semantic segmentation addresses the cross-domain problem with fine source domain labels. However, the acquisition of semantic labels has always been a difficult step, many scenarios only have weak labels (e.g. bounding boxes). For scenarios where weak supervision and cross-domain problems coexist, this paper defines a new task: unsupervised domain adaptation based on weak source domain labels (WUDA). To explore solutions for this task, this paper proposes two intuitive frameworks: 1) Perform weakly supervised semantic segmentation in the source domain, and then implement unsupervised domain adaptation; 2) Train an object detection model using source domain data, then detect objects in the target domain and implement weakly supervised semantic segmentation. We observe that the two frameworks behave differently when the datasets change. Therefore, we construct dataset pairs with a wide range of domain shifts and conduct extended experiments to analyze the impact of different domain shifts on the two frameworks. In addition, to measure domain shift, we apply the metric representation shift to urban landscape image segmentation for the first time. The source code and constructed datasets are available at \url{https://github.com/bupt-ai-cz/WUDA}.

arXiv:2210.01795공식 코드 · Papers with Code

BayesFT: Bayesian Optimization for Fault Tolerant Neural Network Architecture

To deploy deep learning algorithms on resource-limited scenarios, an emerging device-resistive random access memory (ReRAM) has been regarded as promising via analog computing. However, the practicability of ReRAM is primarily limited due to the weight drifting of ReRAM neural networks due to multi-factor reasons, including manufacturing, thermal noises, and etc. In this paper, we propose a novel Bayesian optimization method for fault tolerant neural network architecture (BayesFT). For neural architecture search space design, instead of conducting neural architecture search on the whole feasible neural architecture search space, we first systematically explore the weight drifting tolerance of different neural network components, such as dropout, normalization, number of layers, and activation functions in which dropout is found to be able to improve the neural network robustness to weight drifting. Based on our analysis, we propose an efficient search space by only searching for dropout rates for each layer. Then, we use Bayesian optimization to search for the optimal neural architecture robust to weight drifting. Empirical experiments demonstrate that our algorithmic framework has outperformed the state-of-the-art methods by up to 10 times on various tasks, such as image classification and object detection.

arXiv:2210.00087공식 코드 · Papers with Code

D-Align: Dual Query Co-attention Network for 3D Object Detection Based on Multi-frame Point Cloud Sequence

LiDAR sensors are widely used for 3D object detection in various mobile robotics applications. LiDAR sensors continuously generate point cloud data in real-time. Conventional 3D object detectors detect objects using a set of points acquired over a fixed duration. However, recent studies have shown that the performance of object detection can be further enhanced by utilizing spatio-temporal information obtained from point cloud sequences. In this paper, we propose a new 3D object detector, named D-Align, which can effectively produce strong bird's-eye-view (BEV) features by aligning and aggregating the features obtained from a sequence of point sets. The proposed method includes a novel dual-query co-attention network that uses two types of queries, including target query set (T-QS) and support query set (S-QS), to update the features of target and support frames, respectively. D-Align aligns S-QS to T-QS based on the temporal context features extracted from the adjacent feature maps and then aggregates S-QS with T-QS using a gated attention mechanism. The dual queries are updated through multiple attention layers to progressively enhance the target frame features used to produce the detection results. Our experiments on the nuScenes dataset show that the proposed D-Align method greatly improved the performance of a single frame-based baseline method and significantly outperformed the latest 3D object detectors.

arXiv:2210.00975공식 코드 · Papers with Code

DOTIE -- Detecting Objects through Temporal Isolation of Events using a Spiking Architecture

Vision-based autonomous navigation systems rely on fast and accurate object detection algorithms to avoid obstacles. Algorithms and sensors designed for such systems need to be computationally efficient, due to the limited energy of the hardware used for deployment. Biologically inspired event cameras are a good candidate as a vision sensor for such systems due to their speed, energy efficiency, and robustness to varying lighting conditions. However, traditional computer vision algorithms fail to work on event-based outputs, as they lack photometric features such as light intensity and texture. In this work, we propose a novel technique that utilizes the temporal information inherently present in the events to efficiently detect moving objects. Our technique consists of a lightweight spiking neural architecture that is able to separate events based on the speed of the corresponding objects. These separated events are then further grouped spatially to determine object boundaries. This method of object detection is both asynchronous and robust to camera noise. In addition, it shows good performance in scenarios with events generated by static objects in the background, where existing event-based algorithms fail. We show that by utilizing our architecture, autonomous navigation systems can have minimal latency and energy overheads for performing object detection.

arXiv:2210.00808공식 코드 · Papers with Code

A Multi Camera Unsupervised Domain Adaptation Pipeline for Object Detection in Cultural Sites through Adversarial Learning and Self-Training

Object detection algorithms allow to enable many interesting applications which can be implemented in different devices, such as smartphones and wearable devices. In the context of a cultural site, implementing these algorithms in a wearable device, such as a pair of smart glasses, allow to enable the use of augmented reality (AR) to show extra information about the artworks and enrich the visitors' experience during their tour. However, object detection algorithms require to be trained on many well annotated examples to achieve reasonable results. This brings a major limitation since the annotation process requires human supervision which makes it expensive in terms of time and costs. A possible solution to reduce these costs consist in exploiting tools to automatically generate synthetic labeled images from a 3D model of the site. However, models trained with synthetic data do not generalize on real images acquired in the target scenario in which they are supposed to be used. Furthermore, object detectors should be able to work with different wearable devices or different mobile devices, which makes generalization even harder. In this paper, we present a new dataset collected in a cultural site to study the problem of domain adaptation for object detection in the presence of multiple unlabeled target domains corresponding to different cameras and a labeled source domain obtained considering synthetic images for training purposes. We present a new domain adaptation method which outperforms current state-of-the-art approaches combining the benefits of aligning the domains at the feature and pixel level with a self-training process. We release the dataset at the following link https://iplab.dmi.unict.it/OBJ-MDA/ and the code of the proposed architecture at https://github.com/fpv-iplab/STMDA-RetinaNet.

arXiv:2210.01035공식 코드 · Papers with Code

Expediting Large-Scale Vision Transformer for Dense Prediction without Fine-tuning

Vision transformers have recently achieved competitive results across various vision tasks but still suffer from heavy computation costs when processing a large number of tokens. Many advanced approaches have been developed to reduce the total number of tokens in large-scale vision transformers, especially for image classification tasks. Typically, they select a small group of essential tokens according to their relevance with the class token, then fine-tune the weights of the vision transformer. Such fine-tuning is less practical for dense prediction due to the much heavier computation and GPU memory cost than image classification. In this paper, we focus on a more challenging problem, i.e., accelerating large-scale vision transformers for dense prediction without any additional re-training or fine-tuning. In response to the fact that high-resolution representations are necessary for dense prediction, we present two non-parametric operators, a token clustering layer to decrease the number of tokens and a token reconstruction layer to increase the number of tokens. The following steps are performed to achieve this: (i) we use the token clustering layer to cluster the neighboring tokens together, resulting in low-resolution representations that maintain the spatial structures; (ii) we apply the following transformer layers only to these low-resolution representations or clustered tokens; and (iii) we use the token reconstruction layer to re-create the high-resolution representations from the refined low-resolution representations. The results obtained by our method are promising on five dense prediction tasks, including object detection, semantic segmentation, panoptic segmentation, instance segmentation, and depth estimation.

arXiv:2209.15252공식 코드 · Papers with Code

PointPillars Backbone Type Selection For Fast and Accurate LiDAR Object Detection

3D object detection from LiDAR sensor data is an important topic in the context of autonomous cars and drones. In this paper, we present the results of experiments on the impact of backbone selection of a deep convolutional neural network on detection accuracy and computation speed. We chose the PointPillars network, which is characterised by a simple architecture, high speed, and modularity that allows for easy expansion. During the experiments, we paid particular attention to the change in detection efficiency (measured by the mAP metric) and the total number of multiply-addition operations needed to process one point cloud. We tested 10 different convolutional neural network architectures that are widely used in image-based detection problems. For a backbone like MobilenetV1, we obtained an almost 4x speedup at the cost of a 1.13% decrease in mAP. On the other hand, for CSPDarknet we got an acceleration of more than 1.5x at an increase in mAP of 0.33%. We have thus demonstrated that it is possible to significantly speed up a 3D object detector in LiDAR point clouds with a small decrease in detection efficiency. This result can be used when PointPillars or similar algorithms are implemented in embedded systems, including SoC FPGAs. The code is available at https://github.com/vision-agh/pointpillars\_backbone.

arXiv:2209.15639공식 코드 · Papers with Code

F-VLM: Open-Vocabulary Object Detection upon Frozen Vision and Language Models

We present F-VLM, a simple open-vocabulary object detection method built upon Frozen Vision and Language Models. F-VLM simplifies the current multi-stage training pipeline by eliminating the need for knowledge distillation or detection-tailored pretraining. Surprisingly, we observe that a frozen VLM: 1) retains the locality-sensitive features necessary for detection, and 2) is a strong region classifier. We finetune only the detector head and combine the detector and VLM outputs for each region at inference time. F-VLM shows compelling scaling behavior and achieves +6.5 mask AP improvement over the previous state of the art on novel categories of LVIS open-vocabulary detection benchmark. In addition, we demonstrate very competitive results on COCO open-vocabulary detection benchmark and cross-dataset transfer detection, in addition to significant training speed-up and compute savings. Code will be released at the https://sites.google.com/view/f-vlm/home

arXiv:2209.15159공식 코드 · Papers with Code

MobileViTv3: Mobile-Friendly Vision Transformer with Simple and Effective Fusion of Local, Global and Input Features

MobileViT (MobileViTv1) combines convolutional neural networks (CNNs) and vision transformers (ViTs) to create light-weight models for mobile vision tasks. Though the main MobileViTv1-block helps to achieve competitive state-of-the-art results, the fusion block inside MobileViTv1-block, creates scaling challenges and has a complex learning task. We propose changes to the fusion block that are simple and effective to create MobileViTv3-block, which addresses the scaling and simplifies the learning task. Our proposed MobileViTv3-block used to create MobileViTv3-XXS, XS and S models outperform MobileViTv1 on ImageNet-1k, ADE20K, COCO and PascalVOC2012 datasets. On ImageNet-1K, MobileViTv3-XXS and MobileViTv3-XS surpasses MobileViTv1-XXS and MobileViTv1-XS by 2% and 1.9% respectively. Recently published MobileViTv2 architecture removes fusion block and uses linear complexity transformers to perform better than MobileViTv1. We add our proposed fusion block to MobileViTv2 to create MobileViTv3-0.5, 0.75 and 1.0 models. These new models give better accuracy numbers on ImageNet-1k, ADE20K, COCO and PascalVOC2012 datasets as compared to MobileViTv2. MobileViTv3-0.5 and MobileViTv3-0.75 outperforms MobileViTv2-0.5 and MobileViTv2-0.75 by 2.1% and 1.0% respectively on ImageNet-1K dataset. For segmentation task, MobileViTv3-1.0 achieves 2.07% and 1.1% better mIOU compared to MobileViTv2-1.0 on ADE20K dataset and PascalVOC2012 dataset respectively. Our code and the trained models are available at: https://github.com/micronDLA/MobileViTv3

arXiv:2209.14922공식 코드 · Papers with Code

GDIP: Gated Differentiable Image Processing for Object-Detection in Adverse Conditions

Detecting objects under adverse weather and lighting conditions is crucial for the safe and continuous operation of an autonomous vehicle, and remains an unsolved problem. We present a Gated Differentiable Image Processing (GDIP) block, a domain-agnostic network architecture, which can be plugged into existing object detection networks (e.g., Yolo) and trained end-to-end with adverse condition images such as those captured under fog and low lighting. Our proposed GDIP block learns to enhance images directly through the downstream object detection loss. This is achieved by learning parameters of multiple image pre-processing (IP) techniques that operate concurrently, with their outputs combined using weights learned through a novel gating mechanism. We further improve GDIP through a multi-stage guidance procedure for progressive image enhancement. Finally, trading off accuracy for speed, we propose a variant of GDIP that can be used as a regularizer for training Yolo, which eliminates the need for GDIP-based image enhancement during inference, resulting in higher throughput and plausible real-world deployment. We demonstrate significant improvement in detection performance over several state-of-the-art methods through quantitative and qualitative studies on synthetic datasets such as PascalVOC, and real-world foggy (RTTS) and low-lighting (ExDark) datasets.

arXiv:2209.15001공식 코드 · Papers with Code

Dilated Neighborhood Attention Transformer

Transformers are quickly becoming one of the most heavily applied deep learning architectures across modalities, domains, and tasks. In vision, on top of ongoing efforts into plain transformers, hierarchical transformers have also gained significant attention, thanks to their performance and easy integration into existing frameworks. These models typically employ localized attention mechanisms, such as the sliding-window Neighborhood Attention (NA) or Swin Transformer's Shifted Window Self Attention. While effective at reducing self attention's quadratic complexity, local attention weakens two of the most desirable properties of self attention: long range inter-dependency modeling, and global receptive field. In this paper, we introduce Dilated Neighborhood Attention (DiNA), a natural, flexible and efficient extension to NA that can capture more global context and expand receptive fields exponentially at no additional cost. NA's local attention and DiNA's sparse global attention complement each other, and therefore we introduce Dilated Neighborhood Attention Transformer (DiNAT), a new hierarchical vision transformer built upon both. DiNAT variants enjoy significant improvements over strong baselines such as NAT, Swin, and ConvNeXt. Our large model is faster and ahead of its Swin counterpart by 1.6% box AP in COCO object detection, 1.4% mask AP in COCO instance segmentation, and 1.4% mIoU in ADE20K semantic segmentation. Paired with new frameworks, our large variant is the new state of the art panoptic segmentation model on COCO (58.5 PQ) and ADE20K (49.4 PQ), and instance segmentation model on Cityscapes (45.1 AP) and ADE20K (35.4 AP) (no extra data). It also matches the state of the art specialized semantic segmentation models on ADE20K (58.1 mIoU), and ranks second on Cityscapes (84.5 mIoU) (no extra data).

arXiv:2209.13948공식 코드 · Papers with Code

Obj2Seq: Formatting Objects as Sequences with Class Prompt for Visual Tasks

Visual tasks vary a lot in their output formats and concerned contents, therefore it is hard to process them with an identical structure. One main obstacle lies in the high-dimensional outputs in object-level visual tasks. In this paper, we propose an object-centric vision framework, Obj2Seq. Obj2Seq takes objects as basic units, and regards most object-level visual tasks as sequence generation problems of objects. Therefore, these visual tasks can be decoupled into two steps. First recognize objects of given categories, and then generate a sequence for each of these objects. The definition of the output sequences varies for different tasks, and the model is supervised by matching these sequences with ground-truth targets. Obj2Seq is able to flexibly determine input categories to satisfy customized requirements, and be easily extended to different visual tasks. When experimenting on MS COCO, Obj2Seq achieves 45.7% AP on object detection, 89.0% AP on multi-label classification and 65.0% AP on human pose estimation. These results demonstrate its potential to be generally applied to different visual tasks. Code has been made available at: https://github.com/CASIA-IVA-Lab/Obj2Seq.

arXiv:2209.13933공식 코드 · Papers with Code

DPNet: Dual-Path Network for Real-time Object Detection with Lightweight Attention

The recent advances of compressing high-accuracy convolution neural networks (CNNs) have witnessed remarkable progress for real-time object detection. To accelerate detection speed, lightweight detectors always have few convolution layers using single-path backbone. Single-path architecture, however, involves continuous pooling and downsampling operations, always resulting in coarse and inaccurate feature maps that are disadvantageous to locate objects. On the other hand, due to limited network capacity, recent lightweight networks are often weak in representing large scale visual data. To address these problems, this paper presents a dual-path network, named DPNet, with a lightweight attention scheme for real-time object detection. The dual-path architecture enables us to parallelly extract high-level semantic features and low-level object details. Although DPNet has nearly duplicated shape with respect to single-path detectors, the computational costs and model size are not significantly increased. To enhance representation capability, a lightweight self-correlation module (LSCM) is designed to capture global interactions, with only few computational overheads and network parameters. In neck, LSCM is extended into a lightweight crosscorrelation module (LCCM), capturing mutual dependencies among neighboring scale features. We have conducted exhaustive experiments on MS COCO and Pascal VOC 2007 datasets. The experimental results demonstrate that DPNet achieves state-of the-art trade-off between detection accuracy and implementation efficiency. Specifically, DPNet achieves 30.5% AP on MS COCO test-dev and 81.5% mAP on Pascal VOC 2007 test set, together mwith nearly 2.5M model size, 1.04 GFLOPs, and 164 FPS and 196 FPS for 320 x 320 input images of two datasets.