WELDON: Weakly Supervised Learning of Deep Convolutional Neural Networks

By Thibaut Durand, Nicolas Thomé, Matthieu Cord. 2016

168 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

Not Using the Car to See the Sidewalk — Quantifying and Controlling the Effects of Context in Classification and Segmentation (2018, background): “Classification and segmentation models, built on top of these deep features, can exploit information about object and context to achieve good performance [11, 7, 14].”
ChestX-Ray8: Hospital-Scale Chest X-Ray Database and Benchmarks on Weakly-Supervised Classification and Localization of Common Thorax Diseases (2017, methodology): “4 illustrates the DCNN architecture we adapted, with similarity to several previous weakly-supervised object localization methods [30, 52, 12, 18].”
Multi-evidence Filtering and Fusion for Multi-label Classification, Object Detection and Semantic Segmentation Based on Weakly Supervised Learning (2018, methodology): “Methods in [28, 10, 9] treat the weakly supervised localization problem as an image classification problem, and obtain object locations in specific pooling layers of their networks.”
W-TALC: Weakly-supervised Temporal Activity Localization and Classification (2018, background): “On the other hand, end-to-end learning with categorical loss functions are presented in [15,16,13,47] and recently, the authors in [71] incorporated the proposal generation network in an end-to-end manner.”
Weakly Supervised Complementary Parts Models for Fine-Grained Image Classification From the Bottom Up (2019, methodology): “In [7, 6], the object detection is solved as a classification problem by specific pooling layers in CNNs.”
SIXray: A Large-Scale Security Inspection X-Ray Benchmark for Prohibited Item Discovery in Overlapping Images (2019, background): “This is related to the research in weakly-supervised object localization [3] [6] [33], or a series of work in localizing objects using top-down class activation [8] [7] [40].”
Attribute-Aware Attention Model for Fine-grained Representation Learning (2018, background): “suming and laborious. Moreover, the hand-annotated parts may be suboptimal for image recognition. Another approach is to localize the hidden semantic parts only under the supervision of object labels [9, 28]. Nevertheless, this manner is not reliable enough to mark accurate parts due to weak labels ”
ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis (2020, methodology): “by the community is Global Average Pooling (GAP). In this paper, we investigate the eect of global pooling of spatial features on zero-shot image retrieval. We propose the use of Global KMax Pooling [5] to interpolate between GAP and Global Max Pooling (GMP). Given a convolution feature map of M M d”
Universal Multimodal Representation for Language Understanding (2023, methodology): “[46] by using the simple recurrent unit as text encoder, and the fully convolutional residual ResNet-152 [52] with Weldon pooling [53] as image encoder for our cross-modal retrieval model.”
Weakly Supervised Object Localization and Detection: A Survey (2021, unknown): “Durand-CVPR-2016 [34] None CNN None ImageNet(tag label) CNN Feature extraction network+weakly supervised prediction module General objects”
Decoupling Category-wise Independence and Relevance with Self-attention for Multi-label Image Classification (2019, methodology): “With the development of weakly-supervised learning (WSL), a number of WSL-based methods [11, 12, 13] has obtained promising performance on multi-label image classification.”
WILDCAT: Weakly Supervised Learning of Deep ConvNets for Image Classification, Pointwise Localization and Segmentation (2017, unknown): “WILDCAT also outperforms recent WSL models with different spatial pooling strategies: 17 pt with respect to GAP GoogLeNet [70] which uses a global average pooling and 6 pt with respect to WELDON [13] which uses a kMax+kMin pooling.”
Deep learning in remote sensing: a review (2017, background): “◗ How should problems raised by very limited annotated data in remote sensing be tackled? Is it possible to learn deep hierarchical models for remote-sensing image understanding in a weakly supervised, semisupervised, or even unsupervised way? A few inspiring works in machine learning and computer v”
Global Weighted Average Pooling Bridges Pixel-level Localization and Image-level Classification (2018, methodology): “[31], [32] proposed methods to select multiple high and low scoring regions instead of a single highest scoring region.”
UntrimmedNets for Weakly Supervised Action Recognition and Detection (2017, methodology): “hile those previous architectures all uses the trimmed clips for training. Weakly supervised learning in videos. Weakly supervised learning was extensively studied in object recognition and detection [1,4,9,34], and there were several works adapting this method to learn action models from videos [28”