5 research outputs found
Direct position tracking in GPS using vector correlator
No full textTraditional GPS receivers track the code and carrier of individual satellite signals and estimate the position and velocity of the receiver using trilateration. Different from this two-step approach, direct maximum likelihood estimation of position has proved to be beneficial in weak signal and multipath environments. In this thesis, an architecture of a direct position tracking loop that maximizes an approximation to the maximum likelihood cost function is presented. The unscented Kalman filter is used for direct position tracking using the vector correlator sum. This technique of maximizing the vector correlation sum proves to be beneficial in certain kinds of multipath environments. Further, the tracking loop architecture is validated using experiments with a software receiver.Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2015-09-29 without embargo termsThe student, Athindran Ramesh Kumar, accepted the attached license on 2015-07-20 at 10:34.The student, Athindran Ramesh Kumar, submitted this Thesis for approval on 2015-07-20 at 10:45.This Thesis was approved for publication on 2015-07-21 at 13:30.DSpace SAF Submission Ingestion Package generated from Vireo submission #8557 on 2015-09-29 at 13:23:13Made available in DSpace on 2015-09-29T20:38:43Z (GMT). No. of bitstreams: 3
RAMESHKUMAR-THESIS-2015.pdf: 2871722 bytes, checksum: 7a425af2e3ec7519d8134edcb281288d (MD5)
thesis_source.zip: 44664601 bytes, checksum: 645b132b1bcfe80553dd1c132b75dc78 (MD5)
LICENSE.txt: 4219 bytes, checksum: 4971824cde8ae06bd9006368d7579d9f (MD5)
Previous issue date: 2015-07-2
DiffLoop: Tuning PID controllers by differentiating through the feedback loop
No full textSince most industrial control applications use PID controllers, PID tuning
and anti-windup measures are significant problems. This paper investigates
tuning the feedback gains of a PID controller via back-calculation and
automatic differentiation tools. In particular, we episodically use a cost
function to generate gradients and perform gradient descent to improve
controller performance. We provide a theoretical framework for analyzing this
non-convex optimization and establish a relationship between back-calculation
and disturbance feedback policies. We include numerical experiments on linear
systems with actuator saturation to show the efficacy of this approach.Comment: Extension of paper in 2021 55th Annual Conference on Information
Sciences and Systems (CISS). IEEE, 202
Pack and Detect: Fast Object Detection in Videos Using Region-of-Interest Packing
No full textObject detection in videos is an important task in computer vision for various applications such as object tracking, video summarization and video search. Although great progress has been made in improving the accuracy of object detection in recent years due to the rise of deep neural networks, the state-of-the-art algorithms are highly computationally intensive. In order to address this challenge, we make two important observations in the context of videos: (i) Objects often occupy only a small fraction of the area in each video frame, and (ii) There is a high likelihood of strong temporal correlation between consecutive frames. Based on these observations, we propose Pack and Detect (PaD), an approach to reduce the computational requirements of object detection in videos. In PaD, only selected video frames called anchor frames are processed at full size. In the frames that lie between anchor frames (inter-anchor frames), regions of interest (ROIs) are identified based on the detections in the previous frame. We propose an algorithm to pack the ROIs of each inter-anchor frame together into a reduced-size frame. The computational requirements of the detector are reduced due to the lower size of the input. In order to maintain the accuracy of object detection, the proposed algorithm expands the ROIs greedily to provide additional background around each object to the detector. PaD can use any underlying neural network architecture to process the full-size and reduced-size frames. Experiments using the ImageNet video object detection dataset indicate that PaD can potentially reduce the number of FLOPS required for a frame by . This leads to an overall increase in throughput of on a 2.1 GHz Intel Xeon server with a NVIDIA Titan X GPU at the cost of drop in accuracy.Proceedings of the ACM India Joint International Conference on Data Science and Management of Data. 201
