Differentiable Meta-Learning Model for Few-Shot Semantic Segmentation

Neural networks for fault diagnosis need enough samples for training, but in practical applications, there are often insufficient samples. In order to solve this problem, we propose a wavelet-prototypical network based on fusion of time and frequency domain (WPNF). The time domain and frequency domain information of the vibration signal can be sent to the model simultaneously to expand the characteristics of the data, a parallel two-channel convolutional structure is proposed to process the information of the signal. After that, a wavelet layer is designed to further extract features. Finally, a prototypical layer is applied to train this network. Experimental results show that the proposed method can accurately identify new classes that have never been used during the training phase when the number of samples in each class is very small, and it is far better than other traditional machine learning models in few-shot scenarios.

show abstract

“…

should be trained on the meta-data set

[ 43 ].

is usually collected in advance or extracted from the metadata repositor [ 44 , 45 ]. Meta learning generally learns

by gradient descent method.…”

Section: Related Background and Terminologymentioning

confidence: 99%

Wavelet-Prototypical Network Based on Fusion of Time and Frequency Domain for Fault Diagnosis

Wang

Chen

Liu

et al. 2021

Sensors

View full text Add to dashboard Cite

show abstract

“…Weight imprinting mechanism of new classes using adaptive masked proxies is used in Siam and Oreshkin [6] . CANet [5] uses iterative optimization solution and MetaSegNet [12] uses a differentiable optimization solver for the segmentation task. Recently, prototypical network [8] for fewshot classification is adapted in Wang et al [4] , Dong and Xing [9] to perform the segmentation task.…”

Section: Few-shot Segmentationmentioning

confidence: 99%

“…Training metalearning algorithms constitute two stages of learning: (1) baselearner, which learns to predict an individual task at the episode level and (2) meta-learner, which learns to generalize by learning across a large number of training tasks/episodes. Significant amount of research has been done along these lines, but recently, the approaches based on computing class representatives or class prototypes have been very successful [4,12] .…”

Section: Introductionmentioning

confidence: 99%

SML: Semantic meta-learning for few-shot semantic segmentation☆

Pambala

Dutta

Biswas

2021

Pattern Recognition Letters

View full text Add to dashboard Cite

The significant amount of training data required for training Convolutional Neural Networks has become a bottleneck for applications like semantic segmentation. Few-shot semantic segmentation algorithms address this problem, with an aim to achieve good performance in the low-data regime, with few annotated training images. Recent approaches based on class-prototypes computed from available training data have achieved immense success for this task. In this work, we propose a novel meta-learning framework, Semantic Meta-Learning (SML), which incorporates class level semantic descriptions in the generated prototypes for this problem. In addition, we propose to use the well-established technique, ridge regression, to not only bring in the class-level semantic information, but also to effectively utilise the information available from multiple images present in the training data for prototype computation. This has a simple closed-form solution, and thus can be implemented easily and efficiently. Extensive experiments on the benchmark PASCAL-5i dataset under different experimental settings demonstrate the effectiveness of the proposed framework.

show abstract

“…Additionally, annotating pixel-level labels is often costly in terms of both human efforts and finance, making the current state of the art segmentation models be not suitable for addressing generalized few-shot semantic segmentation problems. Few-shot object segmentation [15,29,35] has received much attention recently due to its advantages in learning novel categories [8,9] without much annotations. Most previous approaches [30,42] follow the metric-based few-shot learning scheme and make great efforts on developing robust feature embedding to measure the pixel-wise similarity between the object from the support image and the query one.…”

Section: Related Workmentioning

confidence: 99%

Meta Parsing Networks: Towards Generalized Few-shot Scene Parsing with Adaptive Metric Learning

Wei

Yang

2020

Proceedings of the 28th ACM International Conference on Multimedia

View full text Add to dashboard Cite

Recent progress in few-shot segmentation usually aims at performing novel object segmentation using a few annotated examples as guidance. In this work, we advance this few-shot segmentation paradigm towards a more challenging yet general scenario, i.e., Generalized Few-shot Scene Parsing (GFSP). In this task, we take a fully annotated image as guidance to segment all pixels in a query image. Our mission is to study a generalizable and robust segmentation network from the meta-learning perspective so that both seen and unseen categories can be correctly recognized. Different from previous practices, this task performs segmentation on a joint label space consisting of both previously seen and novel categories. Moreover, pixels from these multiple categories need to be simultaneously taken into account, which is actually not well explored before. Accordingly, we present Meta Parsing Networks (MPNet) to better exploit the guidance information in the support set. Our MPNet contains two basic modules, i.e., the Adaptive Deep Metric Learning (ADML) module and the Contrastive Inter-class Distraction (CID) module. Specially, the ADML takes the annotated pixels from the support image as the guidance and adaptively produces high-quality prototypes for learning a deep comparison metric. In addition, MPNet further introduces the CID module learning to enlarge the feature discrepancy of different categories in the embedding space, leading the MPNet to generate more discriminative feature embeddings. We conduct experiments on two newly constructed benchmarks, i.e., GFSP-Cityscapes and GFSP-Pascal-Context. Extensive ablation studies well demonstrate the effectiveness and generalization ability of our MPNet. CCS CONCEPTS • Computing methodologies → Computer vision.

show abstract

Differentiable Meta-Learning Model for Few-Shot Semantic Segmentation

Cited by 84 publications

References 14 publications

Wavelet-Prototypical Network Based on Fusion of Time and Frequency Domain for Fault Diagnosis

Wavelet-Prototypical Network Based on Fusion of Time and Frequency Domain for Fault Diagnosis

SML: Semantic meta-learning for few-shot semantic segmentation☆

Meta Parsing Networks: Towards Generalized Few-shot Scene Parsing with Adaptive Metric Learning

Contact Info

Product

Resources

About