TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE 1 ModDrop

By Graham W. Taylor, Florian Nebout. 2018

How researchers see this paper

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

In Optimization & Training

Citations in the corpus

Learning Video Representations from Textual Web Supervision (2020, methodology): “The most commonly-used approach is to use image search results, and to label each image with the query that was used to ﬁnd it [65, 15, 56, 6, 11, 13, 21, 10, 18, 35].”
Geometric Deep Learning on Graphs and Manifolds Using Mixture Model CNNs (2016, methodology): “ls on Euclidean spaces. In this framework, basic ﬁltering operations are represented as convolutions. A signiﬁcant paradigm shift in image processing came with the pioneering work of Perona and Malik [35], suggesting the use of non-shift-invariant image ﬁltering preserving the edge structures. This ”
Pose-conditioned Spatio-Temporal Attention for Human Action Recognition (2017, neutral): “Sequences of frames are stacked into volumes and fed into convolutional layers at first stages [3, 15, 27, 28, 40].”
CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification (2021, background): “, image pyramids [1], scale-space representation [29], and coarse-to-fine approaches [28]).”
Large-Scale Weakly-Supervised Pre-Training for Video Action Recognition (2019, background): “Dataset sources: Many prior approaches use Internet as a natural source of content [6, 12, 13, 18, 24, 25, 34, 37, 44, 46, 51, 58, 67] and the associated search queries, hashtags, or user-comments as labels to construct datasets.”
Domain Adaptation for Visual Applications: A Comprehensive Survey (2017, methodology): “268, 269, 270, 95] or multi-view embedding [93, 110, 99, 111, 112, 113]. Hence, HDA/HTL can strongly beneﬁt from such cross-modal data representations. In the same spirit, webly supervised approaches [271, 272, 273, 274, 158, 159, 275] are also related to DA and HDA as is these approaches rely on co”
GRADE: Gibbs reaction and diffusion equations (1998, background): “Perona and Malik introduced a family of anisotropic diffusion equations for generating image scale space I(z, y, t ) [17].”
An Analysis of Scale Invariance in Object Detection - SNIP (2017, background): “Scale space theory [35, 26] advocates learning representations that are invariant to scale and the theory has been applied to many problems in the history of computer vision [4, 30, 28, 21, 14, 5, 23].”
Web-Based Classifiers for Human Action Recognition (2012, methodology): “These works mostly use a nearest neighbor approach and use global image features or textual information [27].”
Hyperspectral Anomaly Detection With Attribute and Edge-Preserving Filters (2017, background): “The most popular filters in the edge-preserving filtering class are the bilateral filter [30] and anisotropic diffusion [31].”
Community Discovery in Dynamic Networks (2017, background): “…Density (Radicchi et al. , 2004), i.e., the ratio of edges within the cluster w.r.t. all possible edges; • Cut Ratio and Normalized Cut (Fortunato, 2010; Shi and Malik, 2000), i.e., the fraction of all possible edges leaving the cluster; • Maximum/Average ODF (Flake et al. , 2000) (out-degree…”
Algorithm Unrolling: Interpretable, Efficient Deep Learning for Signal and Image Processing (2019, methodology): “In [14], Chen and Pock adopt the unrolling approach to improve the performance of Perone-Malik anisotropic diffusion [57], a well-known technique for image restoration and edge detection.”
Multi-Stage Progressive Image Restoration (2021, unknown): “In order to restrict the solution space to valid/natural images, existing restoration techniques [19, 29, 39, 59, 66, 67, 100] explicitly use image priors that are handcrafted with empirical observations.”