Today, visual recognition systems are still rarely employed in robotics applications. Perhaps one of the main reasons for this is the lack of demanding benchmarks that mimic such scenarios. In this paper, we take advantage of our autonomous driving platform to develop novel challenging benchmarks for the tasks of stereo, optical flow, visual odometry/SLAM and 3D object detection. Our recording platform is equipped with four high resolution video cameras, a Velodyne laser scanner and a state-of-the-art localization system. Our benchmarks comprise 389 stereo and optical flow image pairs, stereo visual odometry sequences of 39.2 km length, and more than 200k 3D object annotations captured in cluttered scenarios (up to 15 cars and 30 pedestrians are visible per image). Results from state-of-the-art algorithms reveal that methods ranking high on established datasets such as Middlebury perform below average when being moved outside the laboratory to the real world. Our goal is to reduce this bias by providing challenging benchmarks with novel difficulties to the computer vision community. Our benchmarks are available online at: www.cvlibs.net/datasets/kitti

/pdf/are-we-ready-for-autonomous-driving-the-kitti-vision-zygefho352.pdf

Are we ready for autonomous driving? The KITTI vision benchmark suite

We investigate architectures of discriminatively trained deep Convolutional Networks (ConvNets) for action recognition in video. The challenge is to capture the complementary information on appearance from still frames and motion between frames. We also aim to generalise the best performing hand-crafted features within a data-driven learning framework.

Our contribution is three-fold. First, we propose a two-stream ConvNet architecture which incorporates spatial and temporal networks. Second, we demonstrate that a ConvNet trained on multi-frame dense optical flow is able to achieve very good performance in spite of limited training data. Finally, we show that multitask learning, applied to two different action classification datasets, can be used to increase the amount of training data and improve the performance on both. Our architecture is trained and evaluated on the standard video actions benchmarks of UCF-101 and HMDB-51, where it is competitive with the state of the art. It also exceeds by a large margin previous attempts to use deep nets for video classification.

https://proceedings.neurips.cc/paper/2014/file/00ec53c4682d36f5c4359f4ae7bd7ba1-Paper.pdf

Two-Stream Convolutional Networks for Action Recognition in Videos

The goal of precipitation nowcasting is to predict the future rainfall intensity in a local region over a relatively short period of time. Very few previous studies have examined this crucial and challenging weather forecasting problem from the machine learning perspective. In this paper, we formulate precipitation nowcasting as a spatiotemporal sequence forecasting problem in which both the input and the prediction target are spatiotemporal sequences. By extending the fully connected LSTM (FC-LSTM) to have convolutional structures in both the input-to-state and state-to-state transitions, we propose the convolutional LSTM (ConvLSTM) and use it to build an end-to-end trainable model for the precipitation nowcasting problem. Experiments show that our ConvLSTM network captures spatiotemporal correlations better and consistently outperforms FC-LSTM and the state-of-the-art operational ROVER algorithm for precipitation nowcasting.

Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting

Models based on deep convolutional networks have dominated recent image interpretation tasks; we investigate whether models which are also recurrent, or “temporally deep”, are effective for tasks involving sequences, visual and otherwise. We develop a novel recurrent convolutional architecture suitable for large-scale visual learning which is end-to-end trainable, and demonstrate the value of these models on benchmark video recognition tasks, image description and retrieval problems, and video narration challenges. In contrast to current models which assume a fixed spatio-temporal receptive field or simple temporal averaging for sequential processing, recurrent convolutional models are “doubly deep” in that they can be compositional in spatial and temporal “layers”. Such models may have advantages when target concepts are complex and/or training data are limited. Learning long-term dependencies is possible when nonlinearities are incorporated into the network state updates. Long-term RNN models are appealing in that they directly can map variable-length inputs (e.g., video frames) to variable length outputs (e.g., natural language text) and can model complex temporal dynamics; yet they can be optimized with backpropagation. Our recurrent long-term models are directly connected to modern visual convnet models and can be jointly trained to simultaneously learn temporal dynamics and convolutional perceptual representations. Our results show such models have distinct advantages over state-of-the-art models for recognition or generation which are separately defined and/or optimized.

/pdf/long-term-recurrent-convolutional-networks-for-visual-3znyib4gwu.pdf

Long-term recurrent convolutional networks for visual recognition and description

Humans perceive the three-dimensional structure of the world with apparent ease. However, despite all of the recent advances in computer vision research, the dream of having a computer interpret an image at the same level as a two-year old remains elusive. Why is computer vision such a challenging problem and what is the current state of the art? Computer Vision: Algorithms and Applications explores the variety of techniques commonly used to analyze and interpret images. It also describes challenging real-world applications where vision is being successfully used, both for specialized applications such as medical imaging, and for fun, consumer-level tasks such as image editing and stitching, which students can apply to their own personal photos and videos. More than just a source of recipes, this exceptionally authoritative and comprehensive textbook/reference also takes a scientific approach to basic vision problems, formulating physical models of the imaging process before inverting them to produce descriptions of a scene. These problems are also analyzed using statistical models and solved using rigorous engineering techniques Topics and features: structured to support active curricula and project-oriented courses, with tips in the Introduction for using the book in a variety of customized courses; presents exercises at the end of each chapter with a heavy emphasis on testing algorithms and containing numerous suggestions for small mid-term projects; provides additional material and more detailed mathematical topics in the Appendices, which cover linear algebra, numerical techniques, and Bayesian estimation theory; suggests additional reading at the end of each chapter, including the latest research in each sub-field, in addition to a full Bibliography at the end of the book; supplies supplementary course material for students at the associated website, http://szeliski.org/Book/. Suitable for an upper-level undergraduate or graduate-level course in computer science or engineering, this textbook focuses on basic techniques that work under real-world conditions and encourages students to push their creative boundaries. Its design and exposition also make it eminently suitable as a unique reference to the fundamental techniques and current research literature in computer vision.

/pdf/computer-vision-algorithms-and-applications-25dn6wu83j.pdf

Computer Vision: Algorithms and Applications

Variational methods are among the most accurate techniques for estimating the optic flow. They yield dense flow fields and can be designed such that they preserve discontinuities, estimate large displacements correctly and perform well under noise and varying illumination. However, such adaptations render the minimisation of the underlying energy functional very expensive in terms of computational costs: Typically one or more large linear or nonlinear equation systems have to be solved in order to obtain the desired solution. Consequently, variational methods are considered to be too slow for real-time performance. In our paper we address this problem in two ways: (i) We present a numerical framework based on bidirectional multigrid methods for accelerating a broad class of variational optic flow methods with different constancy and smoothness assumptions. Thereby, our work focuses particularly on regularisation strategies that preserve discontinuities. (ii) We show by the examples of five classical and two recent variational techniques that real-time performance is possible in all cases--even for very complex optic flow models that offer high accuracy. Experiments show that frame rates up to 63 dense flow fields per second for image sequences of size 160 × 120 can be achieved on a standard PC. Compared to classical iterative methods this constitutes a speedup of two to four orders of magnitude.

/pdf/a-multigrid-platform-for-real-time-motion-computation-with-4tepxqaol6.pdf

A Multigrid Platform for Real-Time Motion Computation with Discontinuity-Preserving Variational Methods

Although variational methods are among the most accurate techniques for estimating the optical flow, they have not yet entered the field of real-time vision Main reason is the great popularity of standard numerical schemes that are easy to implement, however, at the expense of being too slow for real-time performance In our paper we address this problem in two ways: (i) we present an improved version of the highly accurate technique of Brox et al (2004) Thereby we show that a separate robustification of the constancy assumptions is very useful, in particular if the I-norm is used as penalizer As a result, a method is obtained that yields the lowest angular errors in the literature, (ii) We develop an efficient numerical scheme for the proposed approach that allows real-time performance for sequences of size 160 /spl times/ 720 To this end, we combine two hierarchical strategies: a coarse-to-fine warping strategy as implementation of a fixed point iteration for a non-convex optimisation problem and a nonlinear full multigrid method - a so called full approximation scheme (FAS) - for solving the highly nonlinear equation systems at each warping level In the experimental section the advantage of the proposed approach becomes obvious: Outperforming standard numerical schemes by two orders of magnitude frame rates of six high quality flow fields per second are obtained on a 306 GHz Pentium4 PC

https://www.wisdom.weizmann.ac.il/~vision/courses/2006_2/papers/optic_flow_multigrid/bruhn_iccv05.pdf

Towards ultimate motion estimation: combining highest accuracy with real-time performance

We introduce the concept of complementarity between data and smoothness term in modern variational optic flow methods. First we design a sophisticated data term that incorporates HSV colour representation with higher order constancy assumptions, completely separate robust penalisation, and constraint normalisation. Our anisotropic smoothness term reduces smoothing in the data constraint direction instead of the image edge direction, while enforcing a strong filling-in effect orthogonal to it. This allows optimal complementarity between both terms and avoids undesirable interference. The high quality of our complementary optic flow (COF) approach is demonstrated by the current top ranking result at the Middlebury benchmark.

/pdf/complementary-optic-flow-3bkl18y6s1.pdf

Complementary Optic Flow

Since years variational methods belong to the most accurate techniques for computing the optical flow in image sequences However, if based on the grey value constancy assumption only, such techniques are not robust enough to cope with typical illumination changes in real-world data In our paper we tackle this problem in two ways: First we discuss different photometric invariants for the design of illumination-robust variational optical flow methods These invariants are based on colour information and include such concepts as spherical/ conical transforms, normalisation strategies and the differentiation of logarithms Secondly, we embed them into a suitable multichannel generalisation of the highly accurate variational optical flow technique of Brox et al This in turn allows us to access the true potential of such invariants for estimating the optical flow Experiments with synthetic and real-world data demonstrate the success of combining accuracy and robustness: Even under strongly varying illumination, reliable and precise results are obtained

/pdf/illumination-robust-variational-optical-flow-with-1nt0bixlts.pdf

Illumination-robust variational optical flow with photometric invariants

Despite the fact that temporal coherence is undeniably one of the key aspects when processing video data, this concept has hardly been exploited in recent optical flow methods. In this paper, we will present a novel parametrization for multi-frame optical flow computation that naturally enables us to embed the assumption of a temporally coherent spatial flow structure, as well as the assumption that the optical flow is smooth along motion trajectories. While the first assumption is realized by expanding spatial regularization over multiple frames, the second assumption is imposed by two novel first- and second-order trajectorial smoothness terms. With respect to the latter, we investigate an adaptive decision scheme that makes a local (per pixel) or global (per sequence) selection of the most appropriate model possible. Experiments show the clear superiority of our approach when compared to existing strategies for imposing temporal coherence. Moreover, we demonstrate the state-of-the-art performance of our method by achieving Top 3 results at the widely used Middlebury benchmark.

Andrés Bruhn

Papers

A Multigrid Platform for Real-Time Motion Computation with Discontinuity-Preserving Variational Methods

Towards ultimate motion estimation: combining highest accuracy with real-time performance

Complementary Optic Flow

Illumination-robust variational optical flow with photometric invariants

Modeling temporal coherence for optical flow