Visual Word Ambiguity

By Jan van Gemert, Cor J. Veenman, Arnold M. W. Smeulders, Jan-Mark Geusebroek. 2010

861 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

Aggregating Local Image Descriptors into Compact Codes (2012, background): “One among the most popular [27], [28] consists in using soft quantization techniques instead of a k-means.”
Evaluating Color Descriptors for Object and Scene Recognition (2010, background): “For additional details, see Table 4 [41], [ 42 ] and Table 5 [43].”
High-level event recognition in unconstrained videos (2012, neutral): “[41] proposed an interest-”
New trends and ideas in visual concept detection: the MIR flickr retrieval evaluation initiative (2010, methodology): “by soft assignment, [8]), in the choice of classification method, and in the case of SVMs, in the choice of kernel; [32] presents an interesting overview of trade-offs that can be made between accuracy and speed of computation.”
Author manuscript, published in "International Journal of Computer Vision (2013)" International Journal of Computer Vision manuscript No. (will be inserted by the editor) Image Classification with the Fisher Vector: Theory and Practice (2013, neutral): “Several systems are based on the combination of multiple channels corresponding to many different features including (Bergamo and Torresani, 2012, Boiman et al, 2008, Gehler and Nowozin, 2009, VanGemert et al, 2010).”
PatternNet: A Benchmark Dataset for Performance Evaluation of Remote Sensing Image Retrieval (2017, methodology): “retrieval. The extensive experiments demonstrate the superiority of BOVW over these global features. In [31], BOVW is compared with VLAD and its more compact version, product quantized VLAD (VLAD-PQ) [87], for the purpose of geographic image retrieval from satellite imagery. The results show that VL”
Learning mid-level features for recognition (2010, methodology): “[27] (2) SP + soft quantization + kernel SVM − 64.”
Remote Sensing Image Classification: No Features, No Clustering (2015, background): “Vector quantization (hard feature assignment) [15] is used as a baseline for comparison with Fisher vector (FV) [38], kernel codebook encoding (KCE) [22], and LLC [27].”
Empowering Visual Categorization With the GPU (2011, background): “This paper will study how to accelerate the computation of the kernel values on the GPU.”
One shot learning gesture recognition from RGBD images (2012, methodology): “We experiment on depth image using Dollar’s method [8] for STIP detection, HOG3D [10] for cuboid descriptor, kernel codebook [21] for encoding and SVM [4, 6] for classifying BoVW model.”
Improving the Fisher Kernel for Large-Scale Image Classification (2010, methodology): “approaches is to have multiple combinations of patch detectors, descriptors and spatial pyramids (where a combination is often referred to as a “channel”), to train one classifier per channel and then to combine the output of the classifiers [5,6,7,3].”
Ask the locals: Multi-way local pooling for image recognition (2011, methodology): “Recent work has proposed to replace this hard quantization step of individual SIFT descriptors by soft vector quantization [36], or sparse coding [41], including sparse coding of local groups of SIFT descriptors (macro-features) instead of single ones, in our previous work [4].”