Learning Class-Specific Edges for Object Detection and Segmentation

Prasad, Mahesh; Zisserman, Andrew; Fitzgibbon, Andrew; Kumar, M. Pawan; Torr, Philip H. S.

doi:10.1007/11949619_9

Cited by 35 publications

(33 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dollar et al [5] propose an extension of probabilistic boosting trees to combine a rich set of local features. Prasad et al [17] regularize the problem by restricting the set of pixels under consideration to those detected by a lowlevel edge detector, and use simpler local features and a linear SVM classifier. Mairal et al [13] also reason on low-level edges, but learn dictionaries on multiscale RGB patches with sparse coding and use the reconstruction error curves as features for a linear logistic classifier.…”

Section: Related Workmentioning

confidence: 99%

“…In low-level contour detection, the Berkeley Segmentation Data Set (BSDS) [16] and the Precision-Recall methodology of [15] have served that purpose, and were used in [5,13] to evaluate the proposed methods without object-specific knowledge. An alternative approach is to measure the improvement obtained in a given application when switching from low-level to category-specific contours, as was done in [17,13], with an image classification algorithm based on contour matching. However, such a strategy measures only indirectly the accuracy of contours.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Semantic contours from inverse detectors

Hariharan

Pont-Tuset

Bourdev

et al. 2011

2011 International Conference on Computer Vision

1,447

727

View full text Add to dashboard Cite

We study the challenging problem of localizing and classifying category-specific object contours in real world images. For this purpose, we present a simple yet effective method for combining generic object detectors with bottomup contours to identify object contours. We also provide a principled way of combining information from different part detectors and across categories. In order to study the problem and evaluate quantitatively our approach, we present a dataset of semantic exterior boundaries on more than 20, 000 object instances belonging to 20 categories, using the images from the VOC2011 PASCAL challenge [7].

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Semantic contours from inverse detectors

Hariharan

Pont-Tuset

Bourdev

et al. 2011

2011 International Conference on Computer Vision

1,447

727

View full text Add to dashboard Cite

show abstract

“…Salient boundaries are classified against a background by combining local gradient cues, and a probabilistic edge map is produced. The approach described by [31], [1] is perhaps the most relevant to our study. It deals with the task of detecting object boundaries by first selecting candidate boundary pixels extracted by a standard edge detector (e.g.…”

Section: B Object Detection Using Boundary Knowledge-related Workmentioning

confidence: 99%

“…It deals with the task of detecting object boundaries by first selecting candidate boundary pixels extracted by a standard edge detector (e.g. Canny [4]) and then applying the OBJ CUT segmentation algorithm [31], [22].…”

Section: B Object Detection Using Boundary Knowledge-related Workmentioning

confidence: 99%

See 1 more Smart Citation

Automated and Interactive Lesion Detection and Segmentation in Uterine Cervix Images

Alush

Greenspan

Goldberger

2010

IEEE Trans. Med. Imaging

View full text Add to dashboard Cite

Abstract-This paper presents a procedure for automatic extraction and segmentation of a class-specific object (or region) by learning class-specific boundaries. We describe and evaluate the method with a specific focus on the detection of lesion regions in uterine cervix images. The watershed segmentation map of the input image is modeled using an MRF in which watershed regions correspond to binary random variables indicating whether the region is part of the lesion tissue or not. The local pairwise factors on the arcs of the watershed map indicate whether the arc is part of the object boundary. The factors are based on supervised learning of a visual word distribution. The final lesion region segmentation is obtained using a loopy belief propagation applied to the watershed arc-level MRF. Experimental results on real data show state-of-the-art segmentation results on this very challenging task that if necessary, can be interactively enhanced.

show abstract

Cervix type detection using a self‐supervision boosted object detection technique

Bijoy

Akondi

Fathaah

et al. 2022

Int J Imaging Syst Tech

View full text Add to dashboard Cite

Cervical cancer accounts for a large number of fatalities among cancer patients. It is ranked fourth in the total cancer patients and total number of deaths due to cancer. Developing countries account for 70% of the cases and 90% of the fatalities. Contemporary techniques used for screening cervical cancer are PAP smear test and HPV DNA test. Today there are treatments that can successfully prevent cervical cancer if detected at an early stage. Understanding the cervix type is very important for treatment; computational methods can help us classify the cervix type from cervical images. In this study, we propose an ROI proposal network EfficientCenterDet and a self‐supervision boosted training trick that improves the performance of the network with relatively less labeled data. We use 6114 unlabeled images to perform a pretraining task and 1166 labeled images to retrain the ROI proposal network. The proposed model matches the state‐of‐the‐art IOU of FasterRCNN on the ISIC skin lesions dataset while using one‐third of the number of parameters used in FasterRCNN. On MobileODT cervical data, our self‐supervision boosted model achieves 0.632 IOU, a 10% boost over the state‐of‐the‐art FasterRCNN. Introducing an ensembled EfficientNet B4, the cervix type classification stage achieved an accuracy of 87%.

show abstract

Learning Class-Specific Edges for Object Detection and Segmentation

Cited by 35 publications

References 18 publications

Semantic contours from inverse detectors

Semantic contours from inverse detectors

Automated and Interactive Lesion Detection and Segmentation in Uterine Cervix Images

Cervix type detection using a self‐supervision boosted object detection technique

Contact Info

Product

Resources

About