Deformable Part Descriptors for Fine-Grained Recognition and Attribute Prediction

By Ning Zhang, Ryan Farrell, Forrest N. Iandola, Trevor Darrell. 2013

243 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Artificial Intelligence, 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 Aerospace Engineering

1. Jitendra Malik
2. Yuxin Peng
3. Yuanqing Lin
4. Ross B. Girshick
5. Trevor Darrell
6. Vittorio Ferrari
7. Xiangteng He
8. Feng Zhou

In Artificial Intelligence

1. Trevor Darrell
2. Dacheng Tao
3. Li Fei-Fei
4. Alan L. Yuille
5. Yi Yang
6. Quoc V. Le
7. Luc Van Gool
8. Bernt Schiele

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Dacheng Tao
3. Trevor Darrell
4. Xiaogang Wang
5. Ming-Hsuan Yang
6. Yi Yang
7. Jitendra Malik
8. Alan L. Yuille

Citations in the corpus

• DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition (2013, methodology): “In addition, explicitly taking more structured information such as part locations still helps, and provides another significant performance increase, obtaining an accuracy of 64.96%, compared to the 50.98% accuracy reported in (Zhang et al., 2013).”
• CNN Features Off-the-Shelf: An Astounding Baseline for Recognition (2014, methodology): “DPD [44] is a strongly supervised deformable part model using LBP and color features.”
• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <1MB model size (2016, methodology): “However, it has become common practice to apply ImageNet-trained DNN representations to a variety of applications such as fine-grained object recognition [21, 5], logo identification in images [13], and generating sentences about images [6].”
• Learning Deep Features for Discriminative Localization (2015, methodology): “This enables classification-trained CNNs to learn to perform object localization, without using any bounding box annotations.”
• Part-Based R-CNNs for Fine-Grained Category Detection (2014, methodology): “DPD [47] 50.”
• The application of two-level attention models in deep convolutional neural network for fine-grained image classification (2014, methodology): “[25] (DPD) which relies on deformable part based detector [8] to find object and its parts.”
• 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 [1,3,5,7,9,8,14,16,18,20,21,22,28,29,36,37,41,42,49,51,50,66,68,69,71,73,72,76,77,75,78].”
• Deep Image: Scaling up Image Recognition (2015, neutral): “[30] 80.9 Part discovery [44] 81.0 Ours 84.9 GMTL [38] √ √ 44.2 Symbiotic [6] √ √ 59.4 Alignments [17] √ √ 67.0 B-CNN (D,M) [30] √ 81.9 Ours+BBox √ 85.6 PB R-CNN [52] √ √ 73.9 PN-DCN [4] √ √ 75.7 DPD [53] √ √ √ 51.0 POOF [3] √ √ √ 56.8 Nonparametric [18] √ √ √ 57.8 Symbiotic [6] √ √ √ 61.0 DPD+DeCAF”
• PANDA: Pose Aligned Networks for Deep Attribute Modeling (2013, methodology): “DPD and DeCAF We used the publicly available implementations of [27] based on deformable part models and [8] based on CNN trained on ImageNet.”
• Neural Activation Constellations: Unsupervised Part Model Discovery with Convolutional Networks (2015, methodology): “For example, the deformable part descriptor method of [41] uses a supervised version of [15] for training deformable part models, which then allows for pose normalization by comparing corresponding parts.”
• From generic to specific deep representations for visual recognition (2014, neutral): “non-ConvNet [35] [22] [40] [26] [38] [45] [25] [12] [18] [24] [42] [30] [37] [46] [37] [37] [4] 71.”
• Part Detector Discovery in Deep Convolutional Neural Networks (2014, neutral): “For example, [5] relies on deformable part descriptors [17], which is inspired by DPM.”
• Bird Species Categorization Using Pose Normalized Deep Convolutional Nets (2014, methodology): “Among this large body of work, it is a goal of our paper to empirically investigate which methods and techniques are most important toward achieving good performance.”
• Do Semantic Parts Emerge in Convolutional Neural Networks? (2016, background): “The most notable examples of works on semantic part models are fine-grained recognition [8,9,10], generic object detection [5], articulated pose estimation [11,12,13], and attribute prediction [14,15,16].”
• Destruction and Construction Learning for Fine-Grained Image Recognition (2019, background): “Due to the success of deep learning, fine-grained recognition methods have shifted from multistage frameworks based on hand-crafted features [39, 36, 23, 10] to multistage frameworks with CNN features [13, 31, 29].”
• Local Alignments for Fine-Grained Categorization (2014, methodology): “Recent works in computer vision have veri-fied this mechanism (Farrell et al. 2011; Zhang et al. 2012; Chai et al. 2013; Zhang et al. 2013; Berg and Belhumeur 2013).”