Scene recognition and weakly supervised object localization with deformable part-based models

By Megha Pandey, Svetlana Lazebnik. 2011

467 citations recorded by the source corpus.

Subfields: Aerospace Engineering, 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

In Computer Vision and Pattern Recognition

Citations in the corpus

Webly Supervised Learning of Convolutional Networks (2015, methodology): “The outline of our approach is shown in Figure 2.”
CNN Features Off-the-Shelf: An Astounding Baseline for Recognition (2014, background): “26.05 DPM[28] 30.40 Object Bank[23] 37.60 RBow[29] 37.93 BoP[20] 46.10 miSVM[24] 46.40 D-Parts[38] 51.40 IFV[20] 60.77 MLrep[10] 64.03 CNN-SVM 58.44 points are annotated with their ground truth and estimated labels.”
Unsupervised Discovery of Mid-Level Discriminative Patches (2012, methodology): “Further, our simple method by itself outperforms all others that have been tested on this dataset [5,4,37,3].”
Constrained Convolutional Neural Networks for Weakly Supervised Segmentation (2015, background): “There are several heuristics which have been proposed to handle these issues [22, 30, 31].”
Multi-scale Orderless Pooling of Deep Convolutional Activation Features (2014, background): “Figure 5 illustrates such a “scene detection” approach [29,28] on a few SUN images.”
Weakly Supervised Object Localization with Multi-Fold Multiple Instance Learning (2015, neutral): “In addition, Pandey and Lazebnik [35] carefully study how”
Unsupervised object discovery and localization in the wild: Part-based matching with bottom-up region proposals (2015, background): “Typically, weakly-supervised localization [9, 35, 36, 43, 45, 56] requires positive and negative image-level labels for a target object class.”
The State of the Art: Object Retrieval in Paintings using Discriminative Regions (2014, unknown): “DPMs have been trained previously using the entire image as the ROI, but for scene classification [28].”
Learning object class detectors from weakly annotated video (2012, methodology): “However, learning a detector without location annotation is very difficult and performance is still below fully supervised methods [13, 29].”
Building Efficient Deep Hebbian Networks for Image Classification Tasks (2017, methodology): “[13] and Reconfigurable Models (RBoW) [14], as reported in Table 2.”
Weakly Supervised Object Localization with Latent Category Learning (2014, methodology): “, exhaustive search [20, 21], multiple instance learning [13,28,34], inter-intra-class modeling [5,9,23,28,29] and topic model [25,30].”
Unsupervised Feature Learning for RGB-D Based Object Recognition (2012, methodology): “Training linear SVMs on 80 images and testing on 20 images per category on the pre-defined training/test split by the authors, HMP achieves higher accuracy than many state-of-the-art algorithms: spatial pyramid matching (SPM) [28], deformable parts models (DPM) [13], object bank (OB) [24],”
Y Qualification Main task Y Return N N Y User qualified ? N Feedback Pass ? Pass ? Annotator training Annotating images (2017, methodology): “As initialization, in the first iteration we train the classifier using complete images as positive training examples [10, 11, 44, 54, 43, 30].”
Weakly supervised target detection in remote sensing images based on transferred deep features and negative bootstrapping (2015, background): “Finally, the candidate positive set is formed by collecting all the image patches labeled by bounding boxes on the segmented regions (Pandey and Lazebnik 2011).”
Visual and Semantic Knowledge Transfer for Large Scale Semi-Supervised Object Detection (2016, unknown): “bounding box annotations [17], [18], [19], [20], [21], [22],”
Multi-fold MIL Training for Weakly Supervised Object Localization (2014, unknown): “Following [11, 20, 25], during training we discard any images that only contain object instances marked as “difﬁcult” or “truncated”.”
ConceptMap: Mining Noisy Web Data for Concept Learning (2014, unknown): “[44] Kwitt et al.”
Weakly Supervised Learning for Target Detection in Remote Sensing Images (2015, background): “9% of its edge energy as suggested in [15].”