Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

/pdf/deep-residual-learning-for-image-recognition-b75jg61qrq.pdf

Deep Residual Learning for Image Recognition

In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. 
The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

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

In this paper, we address the problem of 3D articulated multi-person tracking in busy street scenes from a moving, human-level observer. In order to handle the complexity of multi-person interactions, we propose to pursue a two-stage strategy. A multi-body detection-based tracker first analyzes the scene and recovers individual pedestrian trajectories, bridging sensor gaps and resolving temporary occlusions. A specialized articulated tracker is then applied to each recovered pedestrian trajectory in parallel to estimate the tracked person's precise body pose over time. This articulated tracker is implemented in a Gaussian Process framework and operates on global pedestrian silhouettes using a learned statistical representation of human body dynamics. We interface the two tracking levels through a guided segmentation stage, which combines traditional bottom-up cues with top-down information from a human detector and the articulated tracker's shape prediction. We show the proposed approach's viability and demonstrate its performance for articulated multi-person tracking on several challenging video sequences of a busy inner-city scenario.

Articulated multi-body tracking under egomotion

This paper presents a multicamera Visual Room (ViRoom). It is constructed from low-cost digital cameras and standard computers running on Linux. Software based synchronized image capture is introduced. A fully automatic self-calibration method for multiple cameras and without any known calibration object is proposed and verified by 3D reconstruction experiments. This handy calibration allows an easy reconfiguration of the setup. Aside from the computers which are usually already available, such a synchronized multicamera setup with six or seven cameras costs less than 1000 USD.

/pdf/viroom-low-cost-synchronized-multicamera-system-and-its-self-25cy3ub6v1.pdf

ViRoom - Low Cost Synchronized Multicamera System and Its Self-calibration

A texture synthesis method is presented that generates similar texture from an example image. It is based on the emulation of simple but rather carefully chosen image intensity statistics. The resulting texture models are compact and no longer require the example image from which they were derived. They make explicit some structural aspects of the textures and the modeling allows knitting together different textures with convincingly looking transition zones. As textures are seldom flat, it is important to also model 3D effects when textures change under changing viewpoint. The simulation of such changes is supported by the model, assuming examples for the different viewpoints are given.

A Compact Model for Viewpoint Dependent Texture Synthesis

We describe an architecture for a multi-camera, multi-resolution surveillance system. The aim is to support a set of distributed static and pan-tilt-zoom (PTZ) cameras and visual tracking algorithms, together with a central supervisor unit. Each camera (and possibly pan-tilt device) has a dedicated process and processor. Asynchronous interprocess communications and archiving of data are achieved in a simple and effective way via a central repository, implemented using an SQL database.

/pdf/a-distributed-camera-system-for-multi-resolution-11py0gmt9l.pdf

A distributed camera system for multi-resolution surveillance

The SVM+ approach has shown excellent performance in visual recognition tasks for exploiting privileged information in the training data. In this paper, we propose two efficient algorithms for solving the linear and kernel SVM+, respectively. For linear SVM+, we absorb the bias term into the weight vector, and formulate a new optimization problem with simpler constraints in the dual form. Then, we develop an efficient dual coordinate descent algorithm to solve the new optimization problem. For kernel SVM+, we further apply the l2-loss, which leads to a simpler optimization problem in the dual form with only half of dual variables when compared with the dual form of the original SVM+ method. More interestingly, we show that our new dual problem can be efficiently solved by using the SMO algorithm of the one-class SVM problem. Comprehensive experiments on three datasets clearly demonstrate that our proposed algorithms achieve significant speed-up than the state-of-the-art solvers for linear and kernel SVM+.

/pdf/fast-algorithms-for-linear-and-kernel-svm-2bkpe2ejxy.pdf

Luc Van Gool

Papers

Articulated multi-body tracking under egomotion

ViRoom - Low Cost Synchronized Multicamera System and Its Self-calibration

A Compact Model for Viewpoint Dependent Texture Synthesis

A distributed camera system for multi-resolution surveillance

Fast Algorithms for Linear and Kernel SVM