Vision is one of the primary sensing modalities in autonomous driving. In this paper we look at the problem of estimating the velocity of road vehicles from a camera mounted on a moving car. Contrary to prior methods that train end-to-end deep networ
ks that estimate the vehicles velocity from the video pixels, we propose a two-step approach where first an off-the-shelf tracker is used to extract vehicle bounding boxes and then a small neural network is used to regress the vehicle velocity from the tracked bounding boxes. Surprisingly, we find that this still achieves state-of-the-art estimation performance with the significant benefit of separating perception from dynamics estimation via a clean, interpretable and verifiable interface which allows us distill the statistics which are crucial for velocity estimation. We show that the latter can be used to easily generate synthetic training data in the space of bounding boxes and use this to improve the performance of our method further.
We present a system for automatic converting of 2D mask object predictions and raw LiDAR point clouds into full 3D bounding boxes of objects. Because the LiDAR point clouds are partial, directly fitting bounding boxes to the point clouds is meaningle
ss. Instead, we suggest that obtaining good results requires sharing information between emph{all} objects in the dataset jointly, over multiple frames. We then make three improvements to the baseline. First, we address ambiguities in predicting the object rotations via direct optimization in this space while still backpropagating rotation prediction through the model. Second, we explicitly model outliers and task the network with learning their typical patterns, thus better discounting them. Third, we enforce temporal consistency when video data is available. With these contributions, our method significantly outperforms previous work despite the fact that those methods use significantly more complex pipelines, 3D models and additional human-annotated external sources of prior information.
In the recent years, many methods demonstrated the ability of neural networks tolearn depth and pose changes in a sequence of images, using only self-supervision as thetraining signal. Whilst the networks achieve good performance, the often over-look
eddetail is that due to the inherent ambiguity of monocular vision they predict depth up to aunknown scaling factor. The scaling factor is then typically obtained from the LiDARground truth at test time, which severely limits practical applications of these methods.In this paper, we show that incorporating prior information about the camera configu-ration and the environment, we can remove the scale ambiguity and predict depth directly,still using the self-supervised formulation and not relying on any additional sensors.
Recent advances in self-supervised learning havedemonstrated that it is possible to learn accurate monoculardepth reconstruction from raw video data, without using any 3Dground truth for supervision. However, in robotics applications,multiple views o
f a scene may or may not be available, depend-ing on the actions of the robot, switching between monocularand multi-view reconstruction. To address this mixed setting,we proposed a new approach that extends any off-the-shelfself-supervised monocular depth reconstruction system to usemore than one image at test time. Our method builds on astandard prior learned to perform monocular reconstruction,but uses self-supervision at test time to further improve thereconstruction accuracy when multiple images are available.When used to update the correct components of the model, thisapproach is highly-effective. On the standard KITTI bench-mark, our self-supervised method consistently outperformsall the previous methods with an average 25% reduction inabsolute error for the three common setups (monocular, stereoand monocular+stereo), and comes very close in accuracy whencompared to the fully-supervised state-of-the-art methods.
We study current-induced deterministic magnetization switching and domain wall motion via polar Kerr microscopy in all-amorphous W$_{66}$Hf$_{34}$/CoFeB/TaO$_text{x}$ with perpendicular magnetic anisotropy and large spin Hall angle. Investigations of
magnetization switching as a function of in-plane assist field and current pulse-width yield switching current densities as low as $3times 10^{9}$ A/m$^2$. We accredit this low switching current density to a low depinning current density, which was obtained from measurements of domain wall displacements upon current injection. This correlation is verified by investigations of a Ta/CoFeB/MgO/Ta reference sample, which showed critical current densities of at least one order of magnitude larger, respectively.
Block coordinate descent (BCD) methods approach optimization problems by performing gradient steps along alternating subgroups of coordinates. This is in contrast to full gradient descent, where a gradient step updates all coordinates simultaneously.
BCD has been demonstrated to accelerate the gradient method in many practical large-scale applications. Despite its success no convergence analysis for inverse problems is known so far. In this paper, we investigate the BCD method for solving linear inverse problems. As main theoretical result, we show that for operators having a particular tensor product form, the BCD method combined with an appropriate stopping criterion yields a convergent regularization method. To illustrate the theory, we perform numerical experiments comparing the BCD and the full gradient descent method for a system of integral equations. We also present numerical tests for a non-linear inverse problem not covered by our theory, namely one-step inversion in multi-spectral X-ray tomography.
Harmonic Hall voltage measurements are a wide-spread quantitative technique for the measurement of spin-orbit induced effective fields in heavy-metal / ferromagnet heterostructures. In the vicinity of the voltage pickup lines in the Hall bar, the cur
rent is inhomogeneous, which leads to a hitherto not quantified reduction of the effective fields and derived quantities, such as the spin Hall angle or the spin Hall conductivity. Here we present a thorough analysis of the influence of the aspect ratio of the voltage pickup lines to current channel widths on the apparent spin Hall angle. Experiments were performed with Hall bars with a broad range of aspect ratios and a substantial reduction of the apparent spin Hall angle is already seen in Hall crosses with an aspect ratio of 1:1. Our experimental results are confirmed by finite-element simulations of the current flow.
The development of accurate and efficient image reconstruction algorithms is a central aspect of quantitative photoacoustic tomography (QPAT). In this paper, we address this issues for multi-source QPAT using the radiative transfer equation (RTE) as
accurate model for light transport. The tissue parameters are jointly reconstructed from the acoustical data measured for each of the applied sources. We develop stochastic proximal gradient methods for multi-source QPAT, which are more efficient than standard proximal gradient methods in which a single iterative update has complexity proportional to the number applies sources. Additionally, we introduce a completely new formulation of QPAT as multilinear (MULL) inverse problem which avoids explicitly solving the RTE. The MULL formulation of QPAT is again addressed with stochastic proximal gradient methods. Numerical results for both approaches are presented. Besides the introduction of stochastic proximal gradient algorithms to QPAT, we consider the new MULL formulation of QPAT as main contribution of this paper.
Quantitative image reconstruction in photoacoustic tomography requires the solution of a coupled physics inverse problem involvier light transport and acoustic wave propagation. In this paper we address this issue employing the radiative transfer equ
ation as accurate model for light transport. As main theoretical results, we derive several stability and uniqueness results for the linearized inverse problem. We consider the case of single illumination as well as the case of multiple illuminations assuming full or partial data. The numerical solution of the linearized problem is much less costly than the solution of the non-linear problem. We present numerical simulations supporting the stability results for the linearized problem and demonstrate that the linearized problem already gives accurate quantitative results.
We investigated the temperature dependence of the switching current for a perpendicularly magnetized CoFeB film deposited on a nanocrystalline tungsten film with large oxygen content: nc-W(O). The spin Hall angle $|Theta_mathrm{SH}| approx 0.22$ is i
ndependent of temperature, whereas the switching current increases strongly at low temperature. We show that the nc-W(O) is insensitive to annealing. It thus can be a good choice for the integration of spin Hall driven writing of information in magnetic memory or logic devices that require a high-temperature annealing process during fabrication.
