Symbiotic Segmentation and Part Localization for Fine-Grained Categorization

By Yuning Chai, Victor S. Lempitsky, Andrew Zisserman. 2013

269 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Fine-Grained Image Classification via Combining Vision and Language (2017, neutral): “Object annotation is used in the works of [12, 13] to learn part detectors in an unsupervised or latent manner.”
Part-Based R-CNNs for Fine-Grained Category Detection (2014, background): “[10] showed that jointly learning part localization and foreground segmentation together can be beneficial for fine-grained categorization.”
The application of two-level attention models in deep convolutional neural network for fine-grained image classification (2014, background): “learn part detectors in unsupervised or latent manner [23, 5].”
The Unreasonable Effectiveness of Noisy Data for Fine-Grained Recognition (2015, background): “The majority of research in fine-grained recognition has focused on developing improved models for classification [2, 5, 7, 15, 17, 18, 24, 30, 37, 44, 45, 58, 59, 60, 63, 64].”
Bilinear CNN Models for Fine-Grained Visual Recognition (2015, methodology): “The next best method is the “symbiotic segmentation” approach of [4] that achieves 72.”
Deep Image: Scaling up Image Recognition (2015, background): “Much work has been devoted on the learning of object parts that benefits from the alignment and segmentation [1, 8, 6, 18].”
Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks (2015, methodology): “[8] use the detections of DPM and the segmentation output of GrabCut to predict part locations.”
The Devil is in the Tails: Fine-grained Classification in the Wild (2017, background): “Most ﬁne-grained research papers present a novel model for classiﬁcation [4, 6, 7, 12, 15, 17, 25, 31, 40, 45, 46, 57, 58, 59, 61, 62].”
Bird Species Categorization Using Pose Normalized Deep Convolutional Nets (2014, methodology): “Areas explored include feature representations that better preserve fine-grained information [35, 46, 47, 48], segmentation-based approaches [1, 13, 14, 15, 21, 37] that facilitate extraction of purer features, and part/pose normalized feature spaces [5, 6, 19, 33, 38, 39, 43, 50, 51].”
Attention for Fine-Grained Categorization (2014, methodology): “As a result, several hand-crafted vision pipelines use provided bounding boxes to isolate the object of interest or may perform segmentation of the object from the background, e.g. Parkhi et al. (2011); Chai et al. (2013); Angelova & Zhu (2013).”
InterActive: Inter-Layer Activeness Propagation (2016, background): “Onthe Bird-200 dataset, it is very important to detect the position and/or compositional parts of the objects [3][13][55][47], otherwise heavy computation is required to achieve good performance [48].”
Augmenting Strong Supervision Using Web Data for Fine-Grained Categorization (2015, methodology): “Various methods have been used to define object parts such as unsupervised patch discovery [32, 8], human-in-the-loop methods [12, 10], or direct reliance on strongly supervised datasets with part-level annotations [21, 4, 33, 20].”
Learning to Navigate for Fine-grained Classification (2018, background): “Some previous works [45,8,3,46,13,2,29] take advantage of fine-grained human annotations, like annotations for bird parts in bird classification.”
Picking Deep Filter Responses for Fine-Grained Image Recognition (2016, methodology): “In order to achieve accurate object and part locations, most existing works explicitly require object level or even part level annotations at both training and testing stages [3], [28], [33].”
Leveraging the Wisdom of the Crowd for Fine-Grained Recognition (2016, unknown): “7 Symbiotic (segmentation only) [51] 39.”
Fine-grained visual categorization via multi-stage metric learning (2014, methodology): “Methods Mean Accuracy (%) Symb [6] 56.”
The Devil is in the Channels: Mutual-Channel Loss for Fine-Grained Image Classification (2020, background): “Some of the earlier works [2], [5], [30], [45] take advantages of bounding-box/part annotations, as an additional information for both training and testing.”