Humans are Still the Best Lossy Image Compressors

Bhown, Ashutosh; Mukherjee, Soham; Yang, Sean; Chandak, Shubham; Fischer-Hwang, Irena; Tatwawadi, Kedar; Weissman, Tsachy

doi:10.1109/dcc.2019.00070

Cited by 4 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Christina Neumayer  https://orcid.org/0000-0003-0450-2983 Notes 1. https://github.com/wondonghyeon/protest-detection-violence-estimation 2. We organized the coders' activity using Amazon Mechanical Turk (AMT) which has become common practice for machine learning research (e.g., Bhown et al, 2019;Hong et al, 2019). To maximize the reliability of the coded data coders were (a) asked to compare two images since pairwise comparison allows the coders to better operate with a complex concept such as violence without forcing them to apply scores (which are prone to be subjective); and (b) we developed a method to rank coders by deviance of their pairwise comparisons from all other coders who coded the same pairs.…”

Section: Orcid Idmentioning

confidence: 99%

Measuring Violence: A Computational Analysis of Violence and Propagation of Image Tweets From Political Protest

Rossi

Neumayer

Henrichsen

et al. 2022

Social Science Computer Review

View full text Add to dashboard Cite

This research quantitatively investigates the impact of violence on the propagation of images in social media in the context of political protest. Using a computational approach, we measure the relative violence of a large set of images shared on Twitter during the protests against the G20 summit in Frankfurt am Main in 2017. This allows us to investigate if more violent content is shared more times and faster than less violent content on Twitter, and if different online communities can be characterized by the level of violence of the visual content they share. The results show that the level of violence in an image tweet does not correlate with the number of retweets and mentions it receives that the time to retweet is marginally lower for image tweets containing a high level of violence and that the level of violence in image tweets differs between communities.

show abstract

Section: Orcid Idmentioning

confidence: 99%

Measuring Violence: A Computational Analysis of Violence and Propagation of Image Tweets From Political Protest

Rossi

Neumayer

Henrichsen

et al. 2022

Social Science Computer Review

View full text Add to dashboard Cite

show abstract

“…Many popular libraries such as MozJPEG still employ use of these specific works [8]. On the other hand, research in mathematical modeling of the Human Visual System (HVS) is still a large ongoing research topic [9,10,11,12].…”

Section: Jpeg Basicsmentioning

confidence: 99%

Modeling Image Quantization Tradeoffs for Optimal Compression

Chiu¹

2021

Preprint

View full text Add to dashboard Cite

All Lossy compression algorithms employ similar compression schemes -frequency domain transform followed by quantization and lossless encoding schemes. They target tradeoffs by quantizating high frequency data to increase compression rates which come at the cost of higher image distortion. We propose a new method of optimizing quantization tables using Deep Learning and a minimax loss function that more accurately measures the tradeoffs between rate and distortion parameters (RD) than previous methods. We design a convolutional neural network (CNN) that learns a mapping between image blocks and quantization tables in an unsupervised manner. By processing images across all channels at once, we can achieve stronger performance by also measuring tradeoffs in information loss between different channels. We initially target optimization on JPEG images but feel that this can be expanded to any lossy compressor.

show abstract

Semantic Image Compression Using Textual Transforms

Arikan,

Weissman

2024

2024 IEEE International Symposium on Information Theory Workshops (ISIT-W)

View full text Add to dashboard Cite

Humans are Still the Best Lossy Image Compressors

Cited by 4 publications

References 0 publications

Measuring Violence: A Computational Analysis of Violence and Propagation of Image Tweets From Political Protest

Measuring Violence: A Computational Analysis of Violence and Propagation of Image Tweets From Political Protest

Modeling Image Quantization Tradeoffs for Optimal Compression

Semantic Image Compression Using Textual Transforms

Contact Info

Product

Resources

About