Oleg Zabluda's blog
Wednesday, October 05, 2016
 
Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks (2013) Ian J.
Multi-digit Number Recognition from Street View Imagery using Deep Convolutional Neural Networks (2013) Ian J. Goodfellow, Yaroslav Bulatov, Julian Ibarz, Sacha Arnoud, Vinay Shet
"""
Traditional approaches to solve this problem typically separate out the localization, segmentation, and recognition steps. In this paper we propose a unified approach that integrates these three steps via the use of a deep convolutional neural network that operates directly on the image pixels. [...] We find that the performance of this approach increases with the depth of the convolutional network, with the best performance occurring in the deepest architecture we trained, with eleven hidden layers. We evaluate this approach on the publicly available SVHN dataset and achieve over 96% accuracy in recognizing complete street numbers. We show that on a per-digit recognition task, we improve upon the state-of-the-art, achieving 97.84% accuracy. We also evaluate this approach on an even more challenging dataset generated from Street View imagery containing several tens of millions of street number annotations and achieve over 90% accuracy. To further explore the applicability of the proposed system to broader text recognition tasks, we apply it to synthetic distorted text from reCAPTCHA. reCAPTCHA is one of the most secure reverse turing tests that uses distorted text to distinguish humans from bots. We report a [superhuman] 99.8% accuracy on the hardest category of reCAPTCHA.
"""
https://arxiv.org/abs/1312.6082
http://static.googleusercontent.com/media/research.google.com/en//pubs/archive/42241.pdf

ICLR14: I Goodfellow: Multi-digit Number Recognition from Street View Imagery using ConvNets [OZ: with Yaroslav Bulatov, Julian Ibarz]
https://www.youtube.com/watch?v=vGPI_JvLoN0
https://arxiv.org/abs/1312.6082

Labels:


- posted by Oleg Zabluda @ 8:36 PM | permalink | | | technorati | bloglines | feedster
| |

Home

Powered by Blogger