Ah丶Weii-CSDN博客

原创 [FGD] Focal and Global Knowledge Distillation for detectors (CVPR. 2022)

.

2022-07-17 22:37:19 1339 1

原创 FreeSOLO: Learning to Segment Objects without Annotations* (CVPR. 2022)

freesolo

2022-07-17 22:27:33 1075 2

原创 DCNet (CVPR. 2021)

1. Motivation这篇文章中，作者针对图1出现的2个问题，指出，由于support 和query img 之间的关系没法完全提取，因为之前的方法都是使用GAP的方法，没有考虑局部信息。其次，对于分类估计错误，以及遮挡的问题，Firstly, relations between support fea- tures and query feature are hardly fully explored in previous few-shot detection works, where gl.

2021-09-22 15:49:11 2029

原创 [OLN] Learning Open-World Object Proposals without Learning to Classify

1. MotivationOur main insight is that the classifiers in existing object proposers or class agnostic detectors impedes such generalization, because the model tends to overfit to labeled objects and treat the unlabeled objects in the training set as ba.

2021-09-06 09:54:34 915

原创 [DeFRCN] Decouple Faster R-CNN for Few-Shot Object Detection(ICCV 2021)

1. Motivation 本文基于Transfer-Learning Based 以及 Faster R-CNN进行改进。本文针对分类和回归任务中存在的矛盾点进行分析：In this paper, we look closely into the conventional Faster R-CNN and analyze its contradictions from two orthogonal perspectives, namely multi-stage (RPN vs. RCNN.

2021-09-02 23:15:11 1397 1

原创 [FSCE]FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding(CVPR. 2021)

1. Motivation本文是基于fine-tuning based方法In this work, we observe and address the essential weakness of the fine- tuning based approach – constantly mislabeling novel in- stances as confusable categories, and improve the few-shot detection performance to t.

2021-08-25 21:48:49 2163

原创 [Retentive R-CNN] Generalized Few-Shot Object Detection without Forgetting(CVPR. 2021)

1. Motivation本文关注于fine-tune后的FSOD模型会在base classes上性能下降的问题。这篇文章构建了Retentive R-CNN，创新点在于Bias-Balance RPN Re-detector，用来在识别novel classes的同时，不降低原有的base classes的精度However, the majority focus merely on the performance of few-shot categories and ignore the ca.

2021-08-18 11:51:10 814

原创 Few-Shot Object Detection via Classification Refinement and Distractor Retreatment(CVPR. 2021)

1. MotivationThe current state-of-the-art approach TFA [17] is still far away from satisfaction compared with those general data-abundant detection tasksGiven the fact that TFA is IOU-aware but less semantic discriminative, our key insight is to enh.

2021-08-09 21:08:58 698

原创 [TFA] Frustratingly Simple Few-Shot Object Detection(ICML. 2020)

1. Contribution 分类任务上的few-shot研究较多，相比之前FSOD收到较少的关注。Detecting rare objects from a few examples is an emerging problem.However, much of this work has focused on basic image classification tasks. In contrast, few-shot object detection has received .

2021-08-09 13:05:57 1457 2

原创 [SRR-FSD] Semantic Relation Reasoning for Shot-Stable Few-Shot Object Detection(CVPR. 2021)

[外链图片转存中…(img-7uQh5Zw9-1628130409148)]1. Motivationfew shot 本身存在的意义：In other words, we are unable to alleviate the situation of scarce cases by simply spend- ing more money on annotation even big data is accessible.Therefore, the study of few-shot

2021-08-07 22:00:24 674

原创 Probablitic two-stage detection

1. MotivationWhile the second stage has a probabilistic interpretation, the combination of the two stages does not.A probabilistic two-stage detector is faster and more accu- rate than both its one- and two-stage precursors.2. ContributionWe build .

2021-08-02 21:06:48 223

原创 [LETR]Line Segment Detection Using Transformers without Edges(CVPR.2021 oral)

1. MotivationDespite its practical and scientific importance, line segment detection remains an unsolved problem in computer vision.Deep learning techniques still consist of heuristics-guided modules such as edge/junction/region detection, line grou.

2021-07-22 15:55:52 477

原创 [OVD]Open-Vocabulary Object Detection Using Captions(CVPR. 2021 oral)

1. MotivationDespite the remarkable accuracy of deep neural networks in object detection, they are costly to train and scale due to supervision requirements.Weakly supervised and zero-shot learning techniques have been explored to scale object detec.

2021-07-21 22:20:08 1120

原创 [Det-AdvProp] Robust and Accuracy Object Detection via Adversarial Learning(CVPR. 2021)

1. Motivation数据增强在分类网络中应用广泛，但是在目标检测中under-explored.Data augmentation has become a de facto component for training high-performance deep image classifiers, but its potential is under-explored for object detection.In this paper, we aim to enhance this l.

2021-07-21 17:21:30 690

原创 [IQDet] (CVPR. 2021)

1. MotivationThe improvements in sampling strategies can be divided into two tendencies.(1) From Static to Dynamic.(2) From Sample-wise to Instance-wise.These sampling strategies might have a few limitations.(1) Static rules are not learnable and.

2021-07-15 22:10:16 326

原创 [Auto-Aug] Scale-aware Automatic Automentation for Object Detection(CVPR. 2021)

1. Motivation这篇文章主要关注于目标检测中的数据增强。This paper focuses on data augmentation for object detection.之前的工作，对于如何将尺度适应性融入网络的方法主要来源与网络的结构（FPN）以及数据增强。Previous work handles this challenge which brings the scale adaptation to the network efficiently mainly f.

2021-07-15 10:03:28 619 1

原创 [OTA]Optimal Transport Assignment for Object Detection(CVPR. 2021)

1. MotivationDeTR [3] examines the idea of global optimal matching. But the Hungarian algo- rithm they adopted can only work in a one-to-one assign- ment manner.One-to-Many 的方法。So far, for the CNN based detectors in one-to-many scenarios, a global .

2021-07-14 15:23:49 2405

原创 python logging日志笔记

import loggingimport oslogging.basicConfig( format='[%(asctime)s] %(message)s', # format 和 datefmt都要有。 datefmt='%Y/%M/%d %H:%M:%S', level=logging.DEBUG, handlers=[ logging.FileHandler(os.path.join('/home/you/you/chenwei/AdelaiDe

2021-07-09 22:34:40 82

原创 Cycle GAN(ICCV. 2017)

1. MotivationFor many tasks, paired training data will not be available.We present an approach for learning to translate an image from a source domain X to a target domain Y in the absence of paired examples.对于Cycle consistent的引入很重要，否则会出现模式崩溃的问题，所有输入.

2021-06-30 20:31:15 122

原创 Leetcode Daily

文章目录342. 4的幂1744. 你能在你最喜欢的那天吃到你最喜欢的糖果吗？523. 连续的子数组和525. 连续数组160. 相交链表203. 移除链表元素[474. 一和零](https://leetcode-cn.com/problems/ones-and-zeroes/)[797. 所有可能的路径](https://leetcode-cn.com/problems/all-paths-from-source-to-target/)[494. 目标和](https://leetcode-cn.com

2021-06-30 20:29:01 195

原创 Pix2Pix GAN(CVPR. 2017)

1. MotivationImage-to-Image translation的定义We define automatic image-to-image translation as the task of translating one possible representation of a scene into another.Our goal in this paper is to develop a common framework for all these problems.需要.

2021-06-28 22:41:34 692

原创 GAN[NIPS. 2014]

KL divergence，如果KL的值越大，代表2个分布之间的差异越大，KL的值跃萧，代表2个分布之间的差异最小。KL divergence:DKL(P∣∣Q)=∑i=1NP(xi)logP(xi)Q(xi)D_{KL}(P||Q) = \sum^N_{i=1} P(x_i)log\frac{P(x_i)}{Q(x_i)}DKL(P∣∣Q)=i=1∑NP(xi)logQ(xi)P(xi)JS divergence:JSD(P∣∣Q)=12D(P∣∣M)+12D(Q∣∣M)M=1..

2021-06-23 17:05:12 102

原创 [detectron2 ] Mask R-CNN代码笔记

RPNRPN中的`match_label`表示的是[0,-1,1]# forward pairwise_iou functionmatch_quality_matrix = retry_if_cuda_oom(pairwise_iou)(gt_boxes_i, anchors) # [M x N] for each img# forward match functionmatched_idxs, gt_labels_i = retry_if_cuda_oom(self.anchor_matche

2021-06-21 11:41:18 2027

原创 [PolarMask++](TPAMI. 2021）

1. ContributionThe main contributions of this work are three-fold:We introduce a new perspective to design a single-shot instance segmentation framework, PolarMask, which predicts instance masks and rotated objects in the polar coordinate in an effecti.

2021-06-15 11:37:33 506

原创 [SOLQ]SOLQ: Segmenting Objects by Learning Queries

1. MotivationAnchor-based segmentation: the segmentation branch heavily relies on the detection branch, making it hard to achieve better joint learning of multiple tasks.Anchor-free segmentation: the weights of dynamic convolutions or the mask coeffic.

2021-06-15 10:54:36 767

原创 [CenterMask]CenterMask: Real-Time Anchor-Free Instance Segmentation(CVPR. 2020)

1. ContributionWe design a scale-adaptive RoI assignment function that considers the input scale and is a more suitable one-stage object detector.We also propose a more effective backbone network VoVNetV2 based on VoVNet, which shows better performa.

2021-06-09 12:15:05 221

原创 [MLAD] Modeling Multi-Lalbel Action Dependencies for Temporal Action Localization(CVPR. 2021 oral)

1. Motivation以往的大部分方法没有显示建模不同action label的关系。Although these works achieve strong multi-label action localization performance, they do not explicitly model the relationships between the different action labels, which can be extremely useful for determ..

2021-06-08 10:46:53 567

原创 DyCo3D (CVPR. 2021)

1. MotivationPrevious top-performing approaches for point cloud instance segmentation involve a bottom-up strategy, which often includes inefficient operations or complex pipelines, such as grouping over-segmented components, introducing additional step.

2021-05-28 10:54:51 867

原创 [PointNet++] (NPIS 2017)

文章目录1. Motivation2. Contribution3. Method3.1 Hierarchical Point Set Feature Learning3.1.2 Sampling layer3.1.3 Grouping layer3.1.4 PointNet layer3.2 Robust Feature Learning under Non-Uniform Sampling Density3.2.1 Multi-scale grouping (MSG)3.2.2 Multi-reso..

2021-05-25 21:38:09 237

原创 [PointNet] PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation(CVPR. 2017)

文章目录1. Motivation2. Introduction and Related Work2.1 Introduction2.2 Related Work3. Contribution4. Method4.1. Properties of Point Sets in RnR^nRn4.2 PointNet Architecture5. Experiment5.1 applications5.1.1 3D Object Classification5.1.2 3D Object Part Segme.

2021-05-25 18:21:43 225

原创 [WSIS] Weakly-supervised Instance Segmentation via Class-agnostic Learning with Salient Images

1. Motivation弱监督实例分割（WSIS）Weakly-supervised instance segmentation (WSIS) is important in computer vision for at least two reasons.Humans have a strong class-agnostic object segmentation ability and can outline boundaries ofunknown objects precisely.

2021-05-19 20:25:26 665

原创 [MOCO v1] Momentum Constrast for Unsupervised Visual Representation Learning(CVPR 2020)

文章目录1. Motivation and Contribution1.1 Motivation1.2 Contribution2. Method2.1 Contrastive Learning as Dictionary Look-up2.2 Momentum Contrast2.3 Relations to previous mechanisms2.4 pseudo code2.5 Pretext Task2.5.1 Technical details.2.5.2 Shuffling BN3. Ex.

2021-05-18 17:44:33 203

原创 [COD] Camouflaged Object Detection(CVPR 2020.oral)

文章目录1. Motivation2. Contribution3. Relation Work3.1 Generic and Salient Object Detection3.2 Camouflaged Object Detection3.2.1 Types of Camouflage3.2.2 COD Formulation3.2.3 Evaluation Metrics.4. Dataset4.1 Professional Annotation4.2 Dataset Features and ...

2021-05-14 11:34:53 792

原创 [ResMLP]ResMLP: Feedforward networks for image claissification with data-efficient training

文章目录1. Contribution2. Summary3. Methods3.1 The overall ResMLP architecture3.2 The Residual Multi-Perceptron Layer3.3 Relationship to the Vision Transformer.3.4 Class-MLP: MLP with class embedding4. Experiment4.1 Comparison with Transformers and convnets ..

2021-05-12 17:05:49 334

原创 [QueryInst]QueryInst: Parallelly Supervised Mask Query fo Instance Segmentation

# 1. MotivationQuery based object detection。Query based object detection frameworks achieve comparable performance with previous state-of-the-art object detectors.How to fully leverage such frameworks to perform instance segmentation remains an open p.

2021-05-10 21:51:18 1135 4

原创 [VarifocalNet] VarifocalNet: An IoU-aware Dense Object Detector (CVPR. 2021oral)

1. Motivation之前的工作，使用分类分数或者结合分类和定位的分数来筛选候选框。Prior work uses the classification score or a combination of classification and predicted localization scores to rank candidates.在检测中的后处理操作中，一般会使用NMS，通过分类分数来对候选框进行排名，然而这会影响检测的性能，作者认为原因在于分类的分数不是总作为衡量bbox定位精.

2021-05-08 20:46:52 412

原创 [Mixer]MLP-Mixer: An all-MLP Architecture forvision

1. MotivationIn this paper we show that while convolutions and attention are both sufficient for good performance, neither of them are necessary.2. ContributionWe propose the MLP-Mixer architecture (or “Mixer” for short), a competitive but conceptua.

2021-05-08 20:46:10 432

原创 [Vscode] 上传Github笔记

1. 初始化Git首先在github上创建一个repository，修改默认的default branch 为master 后续用main会有问题。然后使用终端git(windows10)或者linux 终端。执行以下操作：# 1. cd your uploaded dirgit init # 2. add ignore-filetouch .gitignore# 3. addgit add -A# 4. upload local repositorygit commit -m

2021-04-28 17:19:44 542

原创 [python] zip 和 *使用的进一步理解

zip(*)例子，拆解rs，如果使用zip(rs)则是错误的rs = [[1,2,3],[1,2,3],[1,2,3],[1,2,3]]ra = list(zip(*rs)) # correct# ra[(1, 1, 1, 1), (2, 2, 2, 2), (3, 3, 3, 3)]rb = list(zip(rs)) #wrong#rb[([1, 2, 3],), ([1, 2, 3],), ([1, 2, 3],), ([1, 2, 3],)]*print(*rs) # 把

2021-04-26 13:43:47 68

原创 [MS R-CNN] Mask Scoring R-CNN(CVPR. 2019 oral)

1. MotivationMask Scoring R-CNN 是在Mask R-CNN的基础之上，由于分类的得分scores没法很好的反映mask的quality，例如有些很高的scores 但是mask的quality却比较差，并且随着cls scores的增大，mask quality并不是呈现一种线性的关系。因此将Smask = Scls * Smask_iou，decompose成2个部分，第一个cls的得分可以直接用faster rcnn的检测分类分支，而maskiou作为本文的一个.

2021-04-23 13:09:59 243 1

空空如也

空空如也