CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement

Cheng, Ho Kei; Chung, Jihoon; Tai, Yu-Wing; Tang, Chi-Keung

doi:10.1109/cvpr42600.2020.00891

Cited by 152 publications

(103 citation statements)

References 33 publications

Supporting

Mentioning

103

Contrasting

Order By: Relevance

“…The first segmentation network (at stage I of cascade) is trained to provide the rough localization of nodules, and the second segmentation network (at stage II of cascade) is trained for fine segmentation based on the rough localization. To our knowledge, in some current cascaded segmentation frameworks, the real output (mask or probability map) or pseudo-label output of the first network is generally fed for training the second network, so that the second network gets contextual information [3,4]. But our preliminary experiments show that the provided context information in first network may do not play a significant auxiliary role for refinement of the second network.…”

Section: Cascaded Segmentation Frameworkmentioning

confidence: 96%

Automatic Segmentation and Classification of Thyroid Nodules in Ultrasound Images with Convolutional Neural Networks

Wang

Yuan

et al. 2021

Segmentation, Classification, and Registration of Multi-Modality Medical Imaging Data

View full text Add to dashboard Cite

Ultrasound image plays an important role in the diagnosis of thyroid disease. Accurate segmentation and classification of thyroid nodules are challenging due to their heterogeneous appearance. In this paper, we propose an efficient cascaded segmentation framework and a dual-attention ResNet-based classification network to automatically achieve the accurate segmentation and classification of thyroid nodules, respectively. We evaluate our methods on the training dataset TN-SCUI 2020 Challenge. The 5-fold cross validation results demonstrate that the proposed methods achieve average IoU of 81.43% in segmentation task, and average F1 score of 83.22% in classification task. Finally, our method ranks the first place of segmentation task on the test set through the final online verification. The source code of the proposed methods is available at https://github.com/WAM AWAMA/TNSCUI2020-Seg-Rank1st.

show abstract

Section: Cascaded Segmentation Frameworkmentioning

confidence: 96%

Automatic Segmentation and Classification of Thyroid Nodules in Ultrasound Images with Convolutional Neural Networks

Wang

Yuan

et al. 2021

Segmentation, Classification, and Registration of Multi-Modality Medical Imaging Data

View full text Add to dashboard Cite

show abstract

“…Generic semantic segmentation via multi-task network cascades is proposed based on CascadePSP. It can improve and modify the local boundary of the training data with any resolution and further strengthen the performance of the existing segmentation networks without slight adjustment [38][39][40].…”

Section: Boundary Refinementmentioning

confidence: 99%

Earthquake-Damaged Buildings Detection in Very High-Resolution Remote Sensing Images Based on Object Context and Boundary Enhanced Loss

et al. 2021

View full text Add to dashboard Cite

Fully convolutional networks (FCN) such as UNet and DeepLabv3+ are highly competitive when being applied in the detection of earthquake-damaged buildings in very high-resolution (VHR) remote sensing images. However, existing methods show some drawbacks, including incomplete extraction of different sizes of buildings and inaccurate boundary prediction. It is attributed to a deficiency in the global context-aware and inaccurate correlation mining in the spatial context as well as failure to consider the relative positional relationship between pixels and boundaries. Hence, a detection method for earthquake-damaged buildings based on the object contextual representations (OCR) and boundary enhanced loss (BE loss) was proposed. At first, the OCR module was separately embedded into high-level feature extractions of the two networks DeepLabv3+ and UNet in order to enhance the feature representation; in addition, a novel loss function, that is, BE loss, was designed according to the distance between the pixels and boundaries to force the networks to pay more attention to the learning of the boundary pixels. Finally, two improved networks (including OB-DeepLabv3+ and OB-UNet) were established according to the two strategies. To verify the performance of the proposed method, two benchmark datasets (including YSH and HTI) for detecting earthquake-damaged buildings were constructed according to the post-earthquake images in China and Haiti in 2010, respectively. The experimental results show that both the embedment of the OCR module and application of BE loss contribute to significantly increasing the detection accuracy of earthquake-damaged buildings and the two proposed networks are feasible and effective.

show abstract

“…The ginjinn predict command allows applying a trained model to predict object occurrences for arbitrary images. Instance segmentations can optionally be refined using CascadePSP (Cheng et al, 2020); while slowing down the predictions, this may considerably improve the quality of the object outlines, especially in case of clear object boundaries. To facilitate the further use of the predictions, GinJinn2 provides various output options: 1) visualization of the predictions on the original images, 2) writing a new COCO annotation file, and 3) saving a cropped image and, if applicable, segmentation mask for each predicted object.…”

Section: Object Detection and Instance Segmentationmentioning

confidence: 99%

GinJinn2: Object detection and segmentation for ecology and evolution

Ott

Lautenschlager

2021

Preprint

View full text Add to dashboard Cite

Proper collection and preparation of empirical data still represent one of the most important, but also expensive steps in ecological and evolutionary/systematic research. Modern machine learning approaches, however, have the potential to automate a variety of tasks, which until recently could only be performed manually. Unfortunately, the application of such methods by researchers outside the field is hampered by technical difficulties, some of which, we believe, can be avoided.Here, we present GinJinn2, a user-friendly toolbox for deep learning-based object detection and instance segmentation on image data. Besides providing a convenient command-line interface to existing software libraries, it comprises several additional tools for data handling, pre- and postprocessing, and building advanced analysis pipelines.We demonstrate the application of GinJinn2 for biological purposes using four exemplary analyses, namely the evaluation of seed mixtures, detection of insects on glue traps, segmentation of stomata, and extraction of leaf silhouettes from herbarium specimens.GinJinn2 will enable users with a primary background in biology to apply deep learning-based methods for object detection and segmentation in order to automate feature extraction from image data.

show abstract

CascadePSP: Toward Class-Agnostic and Very High-Resolution Segmentation via Global and Local Refinement

Cited by 152 publications

References 33 publications

Automatic Segmentation and Classification of Thyroid Nodules in Ultrasound Images with Convolutional Neural Networks

Automatic Segmentation and Classification of Thyroid Nodules in Ultrasound Images with Convolutional Neural Networks

Earthquake-Damaged Buildings Detection in Very High-Resolution Remote Sensing Images Based on Object Context and Boundary Enhanced Loss

GinJinn2: Object detection and segmentation for ecology and evolution

Contact Info

Product

Resources

About