2,536 Hits in 5.4 sec

RSA: Randomized Simulation as Augmentation for Robust Human Action Recognition [article]

Yi Zhang, Xinyue Wei, Weichao Qiu, Zihao Xiao, Gregory D. Hager and Alan Yuille
2019 arXiv   pre-print
We generate large-scale synthetic datasets with randomized nuisance factors.  ...  This is in large part due to the explosion of possible variation in video -- including lighting changes, object variation, movement variation, and changes in surrounding context.  ...  The recent advances of action recognition are partially due to the creation of large-scale labeled video datasets [15, 5, 33] .  ... 
arXiv:1912.01180v1 fatcat:4g5bxwil4vcpjhxn3sc4sanwcy

Video-based descriptors for object recognition

Taehee Lee, Stefano Soatto
2011 Image and Vision Computing  
Modules of our system relate to multi-scale feature selection, tracking, local descriptors, and bag-of-features classification, specifically on baseline algorithms [11] [12] [13] [14] .  ...  This is currently under-played in favor of hand-labeled training data, but time can effectively act as a "weak supervisor" in visual recognition, and we attempt to tap on that role.  ...  Acknowledgments This project was supported in part by ARO 56765-CI, ONR N00014-08-1-0414, AFOSR FA9550-09-1-0427. A video demonstration of the system can be seen at  ... 
doi:10.1016/j.imavis.2011.08.003 fatcat:7mex6a6j3ff2pfc3o6w76auehu

Multi-scale features for identifying individuals in large biological databases: an application of pattern recognition technology to the marbled salamander Ambystoma opacum

Lloyd Gamble, Sai Ravela, Kevin McGarigal
2007 Journal of Applied Ecology  
We develop a pattern recognition algorithm and photo-identification method that uses photographs taken in the field to identify individual marbled salamanders ( Ambystoma opacum ), using their dorsal patterns  ...  We develop, test, and apply a pattern recognition algorithm that enables efficient identification of individual marbled salamanders in a database exceeding 1000 images.  ...  Acknowledgements The first two authors contributed equally to this work. We would like to thank B. Compton, A. Richmond, S. Jackson, C. Griffin, S. Melvin, C. Jenkins and B.  ... 
doi:10.1111/j.1365-2664.2007.01368.x fatcat:r4aw2l36cbgirm7dnqr5wc7mty

Is Rotation a Nuisance in Shape Recognition?

Qiuhong Ke, Yi Li
2014 2014 IEEE Conference on Computer Vision and Pattern Recognition  
Rotation in closed contour recognition is a puzzling nuisance in most algorithms.  ...  and 3) how to use rotation unaware local features for rotation aware shape recognition?  ...  Acknowledgement: we would like to thank Prof. David Jacobs and Arijit Biswas from the University of Maryland at College Park for generously providing the Leafsnap dataset.  ... 
doi:10.1109/cvpr.2014.528 dblp:conf/cvpr/KeL14 fatcat:bvvclc4zd5c6pd4jf3nu3rw5ri

A partial least squares framework for speaker recognition

Balaji Vasan Srinivasan, Dmitry N. Zotkin, Ramani Duraiswami
2011 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In this space, a number of approaches to model inter-class separability and nuisance attribute variability have been proposed.  ...  Modern approaches to speaker recognition (verification) operate in a space of "supervectors" created via concatenation of the mean vectors of a Gaussian mixture model (GMM) adapted from a universal background  ...  Accelerating PLS: Despite the success of PLS, its O(N d) computational cost does not scale well for large sample sizes and large number of features.  ... 
doi:10.1109/icassp.2011.5947548 dblp:conf/icassp/SrinivasanZD11 fatcat:dodfty7wjfhv7b4kspkpwkwl4e

Classification Modulo Invariance, With Application to Face Recognition

Andrew M Fraser, Nicolas W Hengartner, Kevin R Vixie, Brendt E Wohlberg
2003 Journal of Computational And Graphical Statistics  
On the face recognition task, a classifier based on our techniques has an error rate that is 20% lower than that of the best algorithm in a reference software distribution.  ...  them to a face recognition task.  ...  Algorithm In describing the algorithms, when we refer to our example, we mean the face recognition work described in Section 4. A.1.  ... 
doi:10.1198/1061860032634 fatcat:z62ta5penze4pj77yswzusakfi

Permutation invariant SVMs

Pannagadatta K. Shivaswamy, Tony Jebara
2006 Proceedings of the 23rd international conference on Machine learning - ICML '06  
Experiments are shown on character recognition, 3D object recognition and various UCI datasets.  ...  This approach induces permutational invariance in the classifier which can then be directly applied to unusual set-based representations of data.  ...  What is lacking in literature, to the best of our knowledge, are large margin discriminative algorithms that factor out nuisance permutations to maximize classification accuracy.  ... 
doi:10.1145/1143844.1143947 dblp:conf/icml/ShivaswamyJ06 fatcat:o5v7ewtiojhnljp4pb5juh54sm

Learning and matching multiscale template descriptors for real-time detection, localization and tracking

Taehee Lee, Stefano Soatto
2011 CVPR 2011  
Each local descriptor aggregates contrast invariant statistics (normalized intensity and gradient orientation) across scales, in a way that enables matching under significant scale variations.  ...  We describe a system to learn an object template from a video stream, and localize and track the corresponding object in live video.  ...  Since the users' goal is to "explore" the object for later recognition, such a purpose is usually reflected in the video containing a fair sample of the nuisance distribution, as we will see in the discussion  ... 
doi:10.1109/cvpr.2011.5995453 dblp:conf/cvpr/LeeS11 fatcat:a7xpn22ojbebhma6tdnwuh7qzy

A Probabilistic Theory of Deep Learning [article]

Ankit B. Patel, Tan Nguyen, Richard G. Baraniuk
2015 arXiv   pre-print
For instance, visual object recognition involves the unknown object position, orientation, and scale in object recognition while speech recognition involves the unknown voice pronunciation, pitch, and  ...  A grand challenge in machine learning is the development of computational algorithms that match or outperform humans in perceptual inference tasks that are complicated by nuisance variation.  ...  A special thanks to Xaq Pitkow whose keen insight, criticisms and detailed feedback on this work have been instrumental in its development.  ... 
arXiv:1504.00641v1 fatcat:dsqdeopvrjdhrhrd6ytmsjhyzq

TI-POOLING: Transformation-Invariant Pooling for Feature Learning in Convolutional Neural Networks

Dmitry Laptev, Nikolay Savinov, Joachim M. Buhmann, Marc Pollefeys
2016 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
This operator is able to efficiently handle prior knowledge on nuisance variations in the data, such as rotation or scale changes.  ...  On the other hand, we formulate features in convolutional neural networks to be transformation-invariant.  ...  This is most probably due to the fact that fewer canonical positions needs to be handled by the learning algorithm.  ... 
doi:10.1109/cvpr.2016.38 dblp:conf/cvpr/LaptevSBP16 fatcat:56yxphtijjbrbofjkeg7wneaea

Toward a Practical Face Recognition System: Robust Alignment and Illumination by Sparse Representation

A. Wagner, J. Wright, A. Ganesh, Zihan Zhou, H. Mobahi, Yi Ma
2012 IEEE Transactions on Pattern Analysis and Machine Intelligence  
This is mostly due to the difficulty of simultaneously handling variations in illumination, image misalignment, and occlusion in the test image.  ...  Many classic and contemporary face recognition algorithms work well on public data sets, but degrade sharply when they are used in a real recognition system.  ...  JW thanks Allen Yang of UC Berkeley EECS and Robert Fossum of UIUC Mathematics for discussions related to this work, and acknowledges support from a Microsoft Fellowship and the Lemelson-Illinois Student  ... 
doi:10.1109/tpami.2011.112 pmid:21646680 fatcat:z4dbgo5axrcepewttalavm3hse

Delving into Robust Object Detection from Unmanned Aerial Vehicles: A Deep Nuisance Disentanglement Approach [article]

Zhenyu Wu, Karthik Suresh, Priya Narayanan, Hongyu Xu, Heesung Kwon, Zhangyang Wang
2020 arXiv   pre-print
Those nuisances constitute a large number of fine-grained domains, across which the detection model has to stay robust.  ...  (NDFT), for the specific challenging problem of object detection in UAV images, achieving a substantial gain in robustness to those nuisances.  ...  In comparison, the abundance of UAV-specific nuisances will cause the resulting UAV-based detection model to operate in a large number of different fine-grained domains.  ... 
arXiv:1908.03856v2 fatcat:d7gx37iz2nbrhiqg27axtl64nu

Training Domain-invariant Object Detector Faster with Feature Replay and Slow Learner [article]

Chaehyeon Lee, Junghoon Seo, Heechul Jung
2021 arXiv   pre-print
Consequently, on a large-scale UAVDT benchmark, it is shown that our framework can reduce the training time of NDFT from 31 hours to 3 hours while still maintaining the performance.  ...  Previously, nuisance disentangled feature transformation (NDFT) was proposed to build domain-invariant feature extractor with with knowledge of nuisance factors.  ...  We conducted vehicle detection on the UAVDT dataset, a large-scale benchmark, and showed that it performs better than baseline, not using nuisance factor predictions, and performs comparably with NDFT.  ... 
arXiv:2105.14693v1 fatcat:fwuztsfb25hrbg2archp2fmmzy


2012 International journal of pattern recognition and artificial intelligence  
The ensuing results have demonstrated that videos possess unique properties that allow both humans and automated systems to perform recognition accurately in difficult viewing conditions.  ...  We also draw connections between the ways in which humans and current algorithms recognize faces.  ...  The opinions, findings, and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of our sponsors.  ... 
doi:10.1142/s0218001412660024 fatcat:xztw7hmpsjacbogyn22axiq4tq

Single-Sample Face Recognition with Image Corruption and Misalignment via Sparse Illumination Transfer

Liansheng Zhuang, Allen Y. Yang, Zihan Zhou, S. Shankar Sastry, Yi Ma
2013 2013 IEEE Conference on Computer Vision and Pattern Recognition  
Single-sample face recognition is one of the most challenging problems in face recognition.  ...  The new algorithm is robust to image misalignment and pixel corruption, and is able to reduce required training images to one sample per class.  ...  We observe that in addition to the wellunderstood image nuisances aforementioned, one of the remaining challenges in face recognition is indeed the small sample set problem.  ... 
doi:10.1109/cvpr.2013.455 dblp:conf/cvpr/ZhuangYZSM13 fatcat:mirhr2iyvbfojji3pt3zd6b7ii
« Previous Showing results 1 — 15 out of 2,536 results