Multilayer neural networks trained with the back-propagation algorithm constitute the best example of a successful gradient based learning technique. Given an appropriate network architecture, gradient-based learning algorithms can be used to synthesize a complex decision surface that can classify high-dimensional patterns, such as handwritten characters, with minimal preprocessing. This paper reviews various methods applied to handwritten character recognition and compares them on a standard handwritten digit recognition task. Convolutional neural networks, which are specifically designed to deal with the variability of 2D shapes, are shown to outperform all other techniques. Real-life document recognition systems are composed of multiple modules including field extraction, segmentation recognition, and language modeling. A new learning paradigm, called graph transformer networks (GTN), allows such multimodule systems to be trained globally using gradient-based methods so as to minimize an overall performance measure. Two systems for online handwriting recognition are described. Experiments demonstrate the advantage of global training, and the flexibility of graph transformer networks. A graph transformer network for reading a bank cheque is also described. It uses convolutional neural network character recognizers combined with global training techniques to provide record accuracy on business and personal cheques. It is deployed commercially and reads several million cheques per day.

Gradient-based learning applied to document recognition

We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

/pdf/going-deeper-with-convolutions-1yobw2o2ds.pdf

Going deeper with convolutions

There is, I think, something ethereal about i —the square root of minus one. I remember first hearing about it at school. It seemed an odd beast at that time—an intruder hovering on the edge of reality.

Usually familiarity dulls this sense of the bizarre, but in the case of i it was the reverse: over the years the sense of its surreal nature intensified. It seemed that it was impossible to write mathematics that described the real world in …

I and i

We study the question of feature sets for robust visual object recognition; adopting linear SVM based human detection as a test case. After reviewing existing edge and gradient based descriptors, we show experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection. We study the influence of each stage of the computation on performance, concluding that fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and high-quality local contrast normalization in overlapping descriptor blocks are all important for good results. The new approach gives near-perfect separation on the original MIT pedestrian database, so we introduce a more challenging dataset containing over 1800 annotated human images with a large range of pose variations and backgrounds.

/pdf/histograms-of-oriented-gradients-for-human-detection-3nteinziuq.pdf

Histograms of oriented gradients for human detection

Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine multiple low-level image features with high-level context. In this paper, we propose a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also present experiments that provide insight into what the network learns, revealing a rich hierarchy of image features. Source code for the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.

/pdf/rich-feature-hierarchies-for-accurate-object-detection-and-22fvjlggyg.pdf

Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation

A content-based 3D neuroradiologic image retrieval system is being developed at the Robotics Institute of CMU. The special characteristics of this system include: 1) directly dealing with multimodal 3D images (MR/CT); 2) image similarity based on anatomical structures of the human brain; 3) combining both visual and collateral information for indexing and retrieval. A testbed has been implemented for using detected salient visual features for indexing and retrieving 3D images.

/pdf/content-based-3d-neuroradiologic-image-indexing-and-3v2fhnh60g.pdf

Content-based 3D Neuroradiologic Image Indexing and Retrieval: Preliminary Results

This work describes two circular microphone arrays and a square microphone array which can be used for sound localization and sound capture. Sound capture by microphone array is achieved by sum and delay beam former (SDBF). A dedicated PCI 128-channel simultaneous input analog-to-digital (AD) board is developed for a 128 ch microphone array with a maximum sampling rate of 22. 7 /spl mu/s/sample. Simulation of sound pressure distribution of 24 and 128 ch circular microphone array and 128 ch square microphone array are shown. Then a 24 ch circular microphone array and a 128 ch square microphone array have been developed. The 24 ch circular microphone array can capture sound from an arbitrary direction. The 128 ch square microphone array can capture sound from a specific point. Both systems are evaluated by using frequency components of the sound. The circular type system can be used on a mobile robot including humanoid robot and square type can be extended towards room coverage type application.

Microphone array for 2D sound localization and capture

We propose a framework to recover projectivedepth based on image homography and discuss itsapplication to scene analysis of video sequences.We describe a robust homography algorithmwhich incorporates contrast/brightness adjustmentand robust estimation into image registration. Wepresent a camera motion solver to obtain the ego-motion and the real/virtual plane position fromhomography. We then apply the Levenburg-Marquardt method to generate a dense projectivedepth map. We also discuss temporal integrationover video sequences. Finally we present theresults of applying the homography-based videoanalysis to motion detection. 1 Introduction Temporal information redundancy of videosequences allows us to use efficient, incrementalmethods which perform temporal integration ofinformation for gradual refinement.Approaches handling 3D scene analysis of videosequences with camera motion can be classifiedinto two categories: algorithms which use 2Dtransformation or model fitting, and algorithmswhich use 3D geometry analysis. Video sequencesof our interest are taken from a moving airborneplatform where the ego-motion is complex and thescene is relatively distant but not necessarily flat;

Homography-Based 3D Scene Analysis of Video Sequences *

In this paper, we propose an algorithm for sustained tracking of humans, where we combine frame-to-frame articulated motion estimation with a per-frame body detection algorithm The proposed approach can automatically recover from tracking error and drift The frame-to-frame motion estimation algorithm replaces traditional dynamic models within a filtering framework Stable and accurate per-frame motion is estimated via an image-gradient based algorithm that solves a linear constrained least squares system The per-frame detector learns appearance of different body parts and dasiasketchespsila expected gradient maps to detect discriminant pose configurations in images The resulting online algorithm is computationally efficient and has been widely tested on a large dataset of sequences of drivers in vehicles It shows stability and sustained accuracy over thousands of frames

/pdf/on-the-sustained-tracking-of-human-motion-47t0iw4xie.pdf

On the sustained tracking of human motion

The authors introduce a system that automatically segments and classifies features in brain MRIs. It takes 22 minutes to segment 144 structures in a 256/spl times/256/spl times/124 voxel image on an SGI computer with three 194 MHz RIOK processors. The accuracy is comparable to manual segmentation, which can take an expert at least 8 months. The process starts with an atlas, a hand segmented and classified MRI of a normal brain. Given a subject's data, the atlas is warped in 3-D using a hierarchical deformable matching algorithm until it closely matches the subject, i.e. the atlas is customized for the subject. The customized atlas contains the segmentation and classification of the subject's anatomical structures. Qualitative and quantitative evaluations of the system's performance show promise for applications in the quantitative study of brain anatomy. The authors have obtained initial results for finding the normal range of variation in the size and symmetry properties of anatomical structures, and for detecting pathologies.

/pdf/quantitative-study-of-brain-anatomy-3clwos93z7.pdf

Takeo Kanade

Papers

Content-based 3D Neuroradiologic Image Indexing and Retrieval: Preliminary Results

Microphone array for 2D sound localization and capture

Homography-Based 3D Scene Analysis of Video Sequences *

On the sustained tracking of human motion

Quantitative study of brain anatomy