Master's Thesis
Refine
Document Type
- Master's Thesis (6) (remove)
Language
- English (6) (remove)
Has Fulltext
- yes (6)
Is part of the Bibliography
- no (6)
Keywords
- AI Safety (1)
- Autonomous Driving (1)
- DNN Robustness (1)
- Deep Learning (1)
- Machine Learning (1)
- Pedestrian Detection (1)
- Reinforcement Learning (1)
- central limit theorem (1)
- deformable model (1)
- fringe tree (1)
Institute
- Informatik und Mathematik (6) (remove)
Recently, Aumüller and Dietzfelbinger proposed a version of a dual-pivot Quicksort, called "Count", which is optimal among dual-pivot versions with respect to the average number of key comparisons required. In this master's thesis we provide further probabilistic analysis of "Count". We derive an exact formula for the average number of swaps needed by "Count" as well as an asymptotic formula for the variance of the number of swaps and a limit law. Also for the number of key comparisons the asymptotic variance and a limit law are identified. We also consider both complexity measures jointly and find their asymptotic correlation.
Autonomous steering of an electric bicycle based on sensor fusion using model predictive control
(2019)
In this thesis a control and steering module for an autonomous bicycle was developed. Based on sensor fusion and model predictive control, the module is able to trace routes autonomously.
The system is developed to run on a Raspberry Pi. An ultrasonic sensor and a 2D Lidar sensor are used for distance measurements. The vehicle’s position is determined by using GPS signals. Additionally, a camera is used to capture pictures for the roadside detection. In order to recognize the road and the position of the vehicle on it, computer vision techniques are used. The captured images are denoised, Canny edge detection is performed and a perspective transformation is applied. Thereafter a sliding window algorithm selects the edges belonging to the roadside and a second order polynomial is fitted to the selected data. Based on this, the road curvature and the lateral position of the vehicle on the road are calculated. The implemented software is thus able to detect straight and curved roads as well as the vehicle’s lateral offset.
A route planning module was implemented to navigate the vehicle from the start to the destination coordinates. This is done by creating an abstract graph of the roads and using Dijkstra’s algorithm to determine the shortest path.
Four MPC controllers were implemented to control the movements of the vehicle. They are based on state space equations derived from the linear single-track vehicle model. This relatively straightforward model makes it possible to predict the vehicle behavior and is efficient to compute. Each controller was built with different parameters for different vehicle speeds to account for the non-linearity of the system. The controllers simulate the future states of the system at each timeslot and select appropriate control signals for steering, throttle and brakes.
In this thesis, all the components of the steering and control module were individually validated. It was established that the each individual component works as expected and certain constraints and accuracy limits were identified. Finally, the closed loop capabilities of the system were assessed using a test vehicle. Despite some limitations imposed by this setup, it was shown that the control module is indeed capable of autonomously navigating a vehicle and avoiding collisions.
Statistical shape models learn to capture the most characteristic geometric variations of anatomical structures given samples from their population. Accordingly, shape models have become an essential tool for many medical applications and are used in, for example, shape generation, reconstruction, and classification tasks. However, established statistical shape models require precomputed dense correspondence between shapes, often lack robustness, and ignore the global surface topology. This thesis presents a novel neural flow-based shape model that does not require any precomputed correspondence. The proposed model relies on continuous flows of a neural ordinary differential equation to model shapes as deformations of a template. To increase the expressivity of the neural flow and disentangle global, low-frequency deformations from the generation of local, high- frequency details, we propose to apply a hierarchy of flows. We evaluate the performance of our model on two anatomical structures, liver, and distal femur. Our model outperforms state-of-the-art methods in providing an expressive and robust shape prior, as indicated by its generalization ability and specificity. More so, we demonstrate the effectiveness of our shape model on shape reconstruction tasks and find anatomically plausible solutions. Finally, we assess the quality of the emerging shape representation in an unsupervised setting and discriminate healthy from pathological shapes.
AI-based computer vision systems play a crucial role in the environment perception for autonomous driving. Although the development of self-driving systems has been pursued for multiple decades, it is only recently that breakthroughs in Deep Neural Networks (DNNs) have led to their widespread application in perception pipelines, which are getting more and more sophisticated. However, with this rising trend comes the need for a systematic safety analysis to evaluate the DNN's behavior in difficult scenarios as well as to identify the various factors that cause misbehavior in such systems. This work aims to deliver a crucial contribution to the lacking literature on the systematic analysis of Performance Limiting Factors (PLFs) for DNNs by investigating the task of pedestrian detection in urban traffic from a monocular camera mounted on an autonomous vehicle. To investigate the common factors that lead to DNN misbehavior, six commonly used state-of-the-art object detection architectures and three detection tasks are studied using a new large-scale synthetic dataset and a smaller real-world dataset for pedestrian detection. The systematic analysis includes 17 factors from the literature and four novel factors that are introduced as part of this work. Each of the 21 factors is assessed based on its influence on the detection performance and whether it can be considered a Performance Limiting Factor (PLF). In order to support the evaluation of the detection performance, a novel and task-oriented Pedestrian Detection Safety Metric (PDSM) is introduced, which is specifically designed to aid in the identification of individual factors that contribute to DNN failure. This work further introduces a training approach for F1-Score maximization whose purpose is to ensure that the DNNs are assessed at their highest performance. Moreover, a new occlusion estimation model is introduced to replace the missing pedestrian occlusion annotations in the real-world dataset. Based on a qualitative analysis of the correlation graphs that visualize the correlation between the PLFs and the detection performance, this study identified 16 of the initial 21 factors as being PLFs for DNNs out of which the entropy, the occlusion ratio, the boundary edge strength, and the bounding box aspect ratio turned out to be most severely affecting the detection performance. The findings of this study highlight some of the most serious shortcomings of current DNNs and pave the way for future research to address these issues.
We give theorems about asymptotic normality of general additive functionals on patricia tries, derived from results on tries. These theorems are applied to show asymptotic normality of the distribution of random fringe trees in patricia tries. Formulas for asymptotic mean and variance are given. The proportion of fringe trees with 𝑘 keys is asymptotically, ignoring oscillations, given by (1−𝜌(𝑘))/(𝐻 +𝐽)𝑘(𝑘−1) with the source entropy 𝐻, an entropy-like constant 𝐽, that is 𝐻 in the binary case, and an exponentially decreasing function 𝜌(𝑘). Another application gives asymptotic normality of the independence number and the number of 𝑘-protected nodes.
Goal-Conditioned Reinforcement Learning (GCRL) is a popular framework for training agents to solve multiple tasks in a single environment. It is cru- cial to train an agent on a diverse set of goals to ensure that it can learn to generalize to unseen downstream goals. Therefore, current algorithms try to learn to reach goals while simultaneously exploring the environment for new ones (Aubret et al., 2021; Mendonca et al., 2021). This creates a form of the prominent exploration-exploitation dilemma. To relieve the pres- sure of a single agent having to optimize for two competing objectives at once, this thesis proposes the novel algorithm family Goal-Conditioned Re- inforcement Learning with Prior Intrinsic Exploration (GC-π), which sep- arates exploration and goal learning into distinct phases. In the first ex- ploration phase, an intrinsically motivated agent explores the environment and collects a rich dataset of states and actions. This dataset is then used to learn a representation space, which acts as the distance metric for the goal- conditioned reward signal. In the final phase, a goal-conditioned policy is trained with the help of the representation space, and its training goals are randomly sampled from the dataset collected during the exploration phase. Multiple variations of these three phases have been extensively evaluated in the classic AntMaze MuJoCo environment (Nachum et al., 2018). The fi- nal results show that the proposed algorithms are able to fully explore the environment and solve all downstream goals while using every dimension of the state space for the goal space. This makes the approach more flexible compared to previous GCRL work, which only ever uses a small subset of the dimensions for the goals (S. Li et al., 2021a; Pong et al., 2020).