We gratefully acknowledge support from
the Simons Foundation and member institutions.


New submissions

[ total of 3 entries: 1-3 ]
[ showing up to 2000 entries per page: fewer | more ]

New submissions for Fri, 25 Sep 20

[1]  arXiv:2009.11455 [pdf, other]
Title: Packet Compressed Sensing Imaging (PCSI): Robust Image Transmission over Noisy Channels
Comments: 15 pages, 3 figures, 3 tables, presented at ARRL/TAPR Digital Communications Conference 2020, for associated software tool see this https URL
Subjects: Image and Video Processing (eess.IV); Multimedia (cs.MM)

Packet Compressed Sensing Imaging (PCSI) is digital unconnected image transmission method resilient to packet loss. The goal is to develop a robust image transmission method that is computationally trivial to transmit (e.g., compatible with low-power 8-bit microcontrollers) and well suited for weak signal environments where packets are likely to be lost. In other image transmission techniques, noise and packet loss leads to parts of the image being distorted or missing. In PCSI, every packet contains random pixel information from the entire image, and each additional packet received (in any order) simply enhances image quality. Satisfactory SSTV resolution (320x240 pixel) images can be received in ~1-2 minutes when transmitted at 1200 baud AFSK, which is on par with analog SSTV transmission time. Image transmission and reception can occur simultaneously on a computer, and multiple images can be received from multiple stations simultaneously - allowing for the creation of "image nets." This paper presents a simple computer application for Windows, Mac, and Linux that implements PCSI transmission and reception on any KISS compatible hardware or software modem on any band and digital mode.

Replacements for Fri, 25 Sep 20

[2]  arXiv:2005.08210 (replaced) [pdf, other]
Title: Multi-Entity and Multi-Enrollment Key Agreement with Correlated Noise
Authors: Onur Günlü
Comments: To appear in the IEEE Transactions on Information Forensics and Security
Subjects: Information Theory (cs.IT); Cryptography and Security (cs.CR); Multimedia (cs.MM); Signal Processing (eess.SP)
[3]  arXiv:2009.07557 (replaced) [pdf, other]
Title: SLGAN: Style- and Latent-guided Generative Adversarial Network for Desirable Makeup Transfer and Removal
Comments: 9 pages, 9 figures
Subjects: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
[ total of 3 entries: 1-3 ]
[ showing up to 2000 entries per page: fewer | more ]

Disable MathJax (What is MathJax?)

Links to: arXiv, form interface, find, cs, recent, 2009, contact, help  (Access key information)