Feature Transformation Network for Few-Shot Learning

Wang, Xiaoyan; Wang, Hongmei; Zhou, Daming

doi:10.1109/access.2021.3065904

Cited by 4 publications

(3 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Attention mechanism classification methods include CAN [13] , SEGA [15] , CTM [16] , MetaOpt [14] , TPNet [11] , MATANet [12] . Local feature classification methods include MLFRNet [17] .…”

Section: Resultsmentioning

confidence: 99%

Robust network based on embedded convolutional block attention module and earth mover’s distance for few-shot learning

Zhang,

Xin,

Zhang

et al. 2023

Third International Conference on Advanced Algorithms and Signal Image Processing (AASIP 2023)

View full text Add to dashboard Cite

Deep neural networks of training require a large quantity of effectively labeled datasets, and models trained with insufficient samples will cause a variety of problems, such as overfitting and poor generalization ability. Few-shot learning has been proposed to settle the problem, and the methods introduced in some papers have achieved impressive results. However, most of the methods have poor robustness due to the under-utilization of the global information and local information of the image. For solving the above problems, an improved classification network is proposed in this paper, which introduces two modifications. On the one hand, Convolutional Block Attention Module (CBAM) is embedded in the network for extracting global features, thus the model can quickly suit various tasks. On the other hand, the image is randomly divided into tiles and represented as a set of local features, the Earth Mover's Distance (EMD) is employed to calculate the distance between embedded sets of two images. Finally, the weighted sum of these two parts' classification errors, is computed as the total classification error of the proposed model. Experiments are conducted on two benchmark datasets: mini-ImageNet and tiered-ImageNet, and the results demonstrate that our proposed method achieves state-of-theart(SOTA) results in the field of few-shot image classification.

show abstract

“…Attention mechanism classification methods include CAN [13] , SEGA [15] , CTM [16] , MetaOpt [14] , TPNet [11] , MATANet [12] . Local feature classification methods include MLFRNet [17] .…”

Section: Resultsmentioning

confidence: 99%

Robust network based on embedded convolutional block attention module and earth mover’s distance for few-shot learning

Zhang,

Xin,

Zhang

et al. 2023

Third International Conference on Advanced Algorithms and Signal Image Processing (AASIP 2023)

View full text Add to dashboard Cite

show abstract

“…Recently, few-shot learning has emerged as a promising approach to address data scarcity challenges in the medical field. Few-shot learning models excel at identifying new categories without the need for retraining, making them particularly suited for tasks where labeled data are limited or expensive to collect [ 13 ]. Among the various techniques for few-shot learning, one notable method is Model-Agnostic Meta-Learning (MAML) [ 14 ].…”

Section: Introductionmentioning

confidence: 99%

Few-Shot Learning for Medical Image Segmentation Using 3D U-Net and Model-Agnostic Meta-Learning (MAML)

Alsaleh,

Albalawi,

Algosaibi

et al. 2024

Diagnostics

View full text Add to dashboard Cite

Deep learning has attained state-of-the-art results in general image segmentation problems; however, it requires a substantial number of annotated images to achieve the desired outcomes. In the medical field, the availability of annotated images is often limited. To address this challenge, few-shot learning techniques have been successfully adapted to rapidly generalize to new tasks with only a few samples, leveraging prior knowledge. In this paper, we employ a gradient-based method known as Model-Agnostic Meta-Learning (MAML) for medical image segmentation. MAML is a meta-learning algorithm that quickly adapts to new tasks by updating a model’s parameters based on a limited set of training samples. Additionally, we use an enhanced 3D U-Net as the foundational network for our models. The enhanced 3D U-Net is a convolutional neural network specifically designed for medical image segmentation. We evaluate our approach on the TotalSegmentator dataset, considering a few annotated images for four tasks: liver, spleen, right kidney, and left kidney. The results demonstrate that our approach facilitates rapid adaptation to new tasks using only a few annotated images. In 10-shot settings, our approach achieved mean dice coefficients of 93.70%, 85.98%, 81.20%, and 89.58% for liver, spleen, right kidney, and left kidney segmentation, respectively. In five-shot sittings, the approach attained mean Dice coefficients of 90.27%, 83.89%, 77.53%, and 87.01% for liver, spleen, right kidney, and left kidney segmentation, respectively. Finally, we assess the effectiveness of our proposed approach on a dataset collected from a local hospital. Employing five-shot sittings, we achieve mean Dice coefficients of 90.62%, 79.86%, 79.87%, and 78.21% for liver, spleen, right kidney, and left kidney segmentation, respectively.

show abstract

“…This paper proposes a classification network of clean and degraded images based on multi-task learning and consistency regularization with the cosine similarity loss [25], [26]. Our proposed method makes a network learn the classification and degradation levels for degraded images as multi-task learning.…”

Section: Introductionmentioning

confidence: 99%

CNN-Based Classification of Degraded Images Without Sacrificing Clean Images

2021

View full text Add to dashboard Cite

Image classification needs to consider image degradations in practice because an image classification network trained with clean images works poorly for degraded images while digital images usually include some degradations such as JPEG compression. To tackle this problem, a common approach is training classification networks on degraded images with various levels of degradation, e.g. various quality factors for JPEG compression. However, the classification networks do not usually have enough accuracy for clean images because the classification networks have been averagely trained for various levels of degradation. This paper aims to construct a classification network of degraded images without sacrificing the performance of clean images. This paper proposes a network to learn the classification of degraded images and degradation levels of degraded images as multi-task learning. Learning the classification of degraded images is the main task of multi-task learning. On the other hand, learning degradation levels is a sub-task of multi-task learning and reinforces the classification ability. In the proposed network, a feature extractor of the classification is trained with image features acquired from a classification network of clean images by using consistency regularization with the cosine similarity loss. The experimental results using different types of degradations, including JPEG compression, Gaussian blur, Gaussian noise, and salt-and-pepper noise, show that our proposed network has enough ability to classify degraded images without sacrificing the performance for clean images.

show abstract

Feature Transformation Network for Few-Shot Learning

Cited by 4 publications

References 32 publications

Robust network based on embedded convolutional block attention module and earth mover’s distance for few-shot learning

Robust network based on embedded convolutional block attention module and earth mover’s distance for few-shot learning

Few-Shot Learning for Medical Image Segmentation Using 3D U-Net and Model-Agnostic Meta-Learning (MAML)

CNN-Based Classification of Degraded Images Without Sacrificing Clean Images

Contact Info

Product

Resources

About