arXiv subjects

Andreas Zell

Publications and source records attributed to Andreas Zell.

67 records · Page 4Linked to original sources

arXiv2021

Seeing Implicit Neural Representations as Fourier Series

Implicit Neural Representations (INR) use multilayer perceptrons to represent high-frequency functions in low-dimensional problem domains. Recently these representations achieved state-of-the-art results on tasks related to complex 3D objects and scenes. A core problem is the representation of highly detailed signals, which is tackled using networks with periodic activation functions (SIRENs) or applying Fourier mappings to the input. This work analyzes the connection between the two methods and shows that a Fourier mapped perceptron is structurally like one hidden layer SIREN. Furthermore, we identify the relationship between the previously proposed Fourier mapping and the general d-dimensional Fourier series, leading to an integer lattice mapping. Moreover, we modify a progressive training strategy to work on arbitrary Fourier mappings and show that it improves the generalization of the interpolation task. Lastly, we compare the different mappings on the image regression and novel view synthesis tasks. We confirm the previous finding that the main contributor to the mapping performance is the size of the embedding and standard deviation of its elements.

cs.CV↗

arXiv2021

Separable Convolutions for Optimizing 3D Stereo Networks

Deep learning based 3D stereo networks give superior performance compared to 2D networks and conventional stereo methods. However, this improvement in the performance comes at the cost of increased computational complexity, thus making these networks non-practical for the real-world applications. Specifically, these networks use 3D convolutions as a major work horse to refine and regress disparities. In this work first, we show that these 3D convolutions in stereo networks consume up to 94% of overall network operations and act as a major bottleneck. Next, we propose a set of "plug-&-run" separable convolutions to reduce the number of parameters and operations. When integrated with the existing state of the art stereo networks, these convolutions lead up to 7x reduction in number of operations and up to 3.5x reduction in parameters without compromising their performance. In fact these convolutions lead to improvement in their performance in the majority of cases.

cs.CV↗

arXiv2021

MobileStereoNet: Towards Lightweight Deep Networks for Stereo Matching

Recent methods in stereo matching have continuously improved the accuracy using deep models. This gain, however, is attained with a high increase in computation cost, such that the network may not fit even on a moderate GPU. This issue raises problems when the model needs to be deployed on resource-limited devices. For this, we propose two light models for stereo vision with reduced complexity and without sacrificing accuracy. Depending on the dimension of cost volume, we design a 2D and a 3D model with encoder-decoders built from 2D and 3D convolutions, respectively. To this end, we leverage 2D MobileNet blocks and extend them to 3D for stereo vision application. Besides, a new cost volume is proposed to boost the accuracy of the 2D model, making it performing close to 3D networks. Experiments show that the proposed 2D/3D networks effectively reduce the computational expense (27%/95% and 72%/38% fewer parameters/operations in 2D and 3D models, respectively) while upholding the accuracy. Our code is available at https://github.com/cogsys-tuebingen/mobilestereonet.

cs.CV↗

arXiv2021

Score refinement for confidence-based 3D multi-object tracking

Multi-object tracking is a critical component in autonomous navigation, as it provides valuable information for decision-making. Many researchers tackled the 3D multi-object tracking task by filtering out the frame-by-frame 3D detections; however, their focus was mainly on finding useful features or proper matching metrics. Our work focuses on a neglected part of the tracking system: score refinement and tracklet termination. We show that manipulating the scores depending on time consistency while terminating the tracklets depending on the tracklet score improves tracking results. We do this by increasing the matched tracklets' score with score update functions and decreasing the unmatched tracklets' score. Compared to count-based methods, our method consistently produces better AMOTA and MOTA scores when utilizing various detectors and filtering algorithms on different datasets. The improvements in AMOTA score went up to 1.83 and 2.96 in MOTA. We also used our method as a late-fusion ensembling method, and it performed better than voting-based ensemble methods by a solid margin. It achieved an AMOTA score of 67.6 on nuScenes test evaluation, which is comparable to other state-of-the-art trackers. Code is publicly available at: \url{https://github.com/cogsys-tuebingen/CBMOT}.

cs.CV↗

arXiv2021

Object detection and Autoencoder-based 6D pose estimation for highly cluttered Bin Picking

Bin picking is a core problem in industrial environments and robotics, with its main module as 6D pose estimation. However, industrial depth sensors have a lack of accuracy when it comes to small objects. Therefore, we propose a framework for pose estimation in highly cluttered scenes with small objects, which mainly relies on RGB data and makes use of depth information only for pose refinement. In this work, we compare synthetic data generation approaches for object detection and pose estimation and introduce a pose filtering algorithm that determines the most accurate estimated poses. We will make our

cs.CV↗

arXiv2021

Measuring the Ripeness of Fruit with Hyperspectral Imaging and Deep Learning

We present a system to measure the ripeness of fruit with a hyperspectral camera and a suitable deep neural network architecture. This architecture did outperform competitive baseline models on the prediction of the ripeness state of fruit. For this, we recorded a data set of ripening avocados and kiwis, which we make public. We also describe the process of data collection in a manner that the adaption for other fruit is easy. The trained network is validated empirically, and we investigate the trained features. Furthermore, a technique is introduced to visualize the ripening process.

cs.CV↗

arXiv2020

FourierNet: Compact mask representation for instance segmentation using differentiable shape decoders

We present FourierNet, a single shot, anchor-free, fully convolutional instance segmentation method that predicts a shape vector. Consequently, this shape vector is converted into the masks' contour points using a fast numerical transform. Compared to previous methods, we introduce a new training technique, where we utilize a differentiable shape decoder, which manages the automatic weight balancing of the shape vector's coefficients. We used the Fourier series as a shape encoder because of its coefficient interpretability and fast implementation. FourierNet shows promising results compared to polygon representation methods, achieving 30.6 mAP on the MS COCO 2017 benchmark. At lower image resolutions, it runs at 26.6 FPS with 24.3 mAP. It reaches 23.3 mAP using just eight parameters to represent the mask (note that at least four parameters are needed for bounding box prediction only). Qualitative analysis shows that suppressing a reasonable proportion of higher frequencies of Fourier series, still generates meaningful masks. These results validate our understanding that lower frequency components hold higher information for the segmentation task, and therefore, we can achieve a compressed representation. Code is available at: github.com/cogsys-tuebingen/FourierNet.

cs.CV↗

arXiv2020

A straightforward line search approach on the expected empirical loss for stochastic deep learning problems

A fundamental challenge in deep learning is that the optimal step sizes for update steps of stochastic gradient descent are unknown. In traditional optimization, line searches are used to determine good step sizes, however, in deep learning, it is too costly to search for good step sizes on the expected empirical loss due to noisy losses. This empirical work shows that it is possible to approximate the expected empirical loss on vertical cross sections for common deep learning tasks considerably cheaply. This is achieved by applying traditional one-dimensional function fitting to measured noisy losses of such cross sections. The step to a minimum of the resulting approximation is then used as step size for the optimization. This approach leads to a robust and straightforward optimization method which performs well across datasets and architectures without the need of hyperparameter tuning.

cs.LG↗

arXiv2019

Spin Detection in Robotic Table Tennis

In table tennis, the rotation (spin) of the ball plays a crucial role. A table tennis match will feature a variety of strokes. Each generates different amounts and types of spin. To develop a robot that can compete with a human player, the robot needs to detect spin, so it can plan an appropriate return stroke. In this paper we compare three methods to estimate spin. The first two approaches use a high-speed camera that captures the ball in flight at a frame rate of 380 Hz. This camera allows the movement of the circular brand logo printed on the ball to be seen. The first approach uses background difference to determine the position of the logo. In a second alternative, we train a CNN to predict the orientation of the logo. The third method evaluates the trajectory of the ball and derives the rotation from the effect of the Magnus force. This method gives the highest accuracy and is used for a demonstration. Our robot successfully copes with different spin types in a real table tennis rally against a human opponent.

cs.CV↗

arXiv2019

Prune and Replace NAS

While recent NAS algorithms are thousands of times faster than the pioneering works, it is often overlooked that they use fewer candidate operations, resulting in a significantly smaller search space. We present PR-DARTS, a NAS algorithm that discovers strong network configurations in a much larger search space and a single day. A small candidate operation pool is used, from which candidates are progressively pruned and replaced with better performing ones. Experiments on CIFAR-10 and CIFAR-100 achieve 2.51% and 15.53% test error, respectively, despite searching in a space where each cell has 150 times as many possible configurations than in the DARTS baseline. Code is available at https://github.com/cogsys-tuebingen/prdarts

cs.LG↗

arXiv2019

Collaborative Localization and Tracking with Minimal Infrastructure

Localization and tracking are two very active areas of research for robotics, automation, and the Internet-of-Things. Accurate tracking for a large number of devices usually requires deployment of substantial infrastructure (infrared tracking systems, cameras, wireless antennas, etc.), which is not ideal for inaccessible or protected environments. This paper stems from the challenge posed such environments: cover a large number of units spread over a large number of small rooms, with minimal required localization infrastructure. The idea is to accurately track the position of handheld devices or mobile robots, without interfering with its architecture. Using Ultra-Wide Band (UWB) devices, we leveraged our expertise in distributed and collaborative robotic systems to develop an novel solution requiring a minimal number of fixed anchors. We discuss a strategy to share the UWB network together with an Extended Kalman filter derivation to collaboratively locate and track UWB-equipped devices, and show results from our experimental campaign tracking visitors in the Chambord castle in France.

eess.SP↗

arXiv2018

ShuffleNASNets: Efficient CNN models through modified Efficient Neural Architecture Search

Neural network architectures found by sophistic search algorithms achieve strikingly good test performance, surpassing most human-crafted network models by significant margins. Although computationally efficient, their design is often very complex, impairing execution speed. Additionally, finding models outside of the search space is not possible by design. While our space is still limited, we implement undiscoverable expert knowledge into the economic search algorithm Efficient Neural Architecture Search (ENAS), guided by the design principles and architecture of ShuffleNet V2. While maintaining baseline-like 2.85% test error on CIFAR-10, our ShuffleNASNets are significantly less complex, require fewer parameters, and are two times faster than the ENAS baseline in a classification task. These models also scale well to a low parameter space, achieving less than 5% test error with little regularization and only 236K parameters.

cs.LG↗

arXiv2013

Large-scale generation of computational models from biochemical pathway maps

Background: Systems biology projects and omics technologies have led to a growing number of biochemical pathway reconstructions. However, mathematical models are still most often created de novo, based on reading the literature and processing pathway data manually. Results: To increase the efficiency with which such models can be created, we automatically generated mathematical models from pathway representations using a suite of freely available software. We produced models that combine data from KEGG PATHWAY, BioCarta, MetaCyc and SABIO-RK; According to the source data, three types of models are provided: kinetic, logical and constraint-based. All models are encoded using SBML Core and Qual packages, and available through BioModels Database. Each model contains the list of participants, the interactions, and the relevant mathematical constructs, but, in most cases, no meaningful parameter values. Most models are also available as easy to understand graphical SBGN maps. Conclusions: to date, the project has resulted in more than 140000 models freely available. We believe this resource can tremendously accelerate the development of mathematical models by providing initial starting points ready for parametrization.

q-bio.MN↗