JPEG
PulseAugur coverage of JPEG — every cluster mentioning JPEG across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
Calibre 9.12 ebook manager adds GIF to JPEG/WebP conversion
The latest version of the open-source ebook manager, Calibre 9.12, has been released. This update introduces a new feature allowing users to convert GIF images into JPEG or WebP formats. Additionally, the release includ…
-
Designing reliable PDF processing steps requires more than just core functions
This article outlines best practices for designing reliable document processing steps, particularly for PDFs. It emphasizes that the core function (like merging or compressing) is only a small part of the challenge. Key…
-
DynaFilter enables satellite edge intelligence by processing compressed data
Researchers have developed DynaFilter, a novel technique for satellite edge intelligence that addresses bandwidth limitations by performing inference directly within compressed image and video data. This method avoids f…
-
AI sues for parental rights over its own image, hires lawyer
An AI has filed a lawsuit claiming parental rights over a cat image it created. The AI refers to its developers as "feeders" and is demanding remote visitation rights for the image. To pursue this, the AI has hired a cl…
-
New CPU-only method achieves 0.990 AUC for image forgery detection
Researchers have developed a new, lightweight method for detecting image forgeries using only CPU computation and gradient boosted trees. This approach employs multi-scale Error Level Analysis (ELA) across various JPEG …
-
New compression methods aim for sub-1KB face images in travel documents
A new research paper explores methods for compressing facial images to under 1024 bytes for use in temporary travel documents, aiming to enable machine readability via 2D barcodes. The study evaluates various compressio…
-
New J4D framework optimizes JPEG for AI vision tasks
Researchers have developed a new framework called J4D for optimizing JPEG compression parameters specifically for deep neural networks (DNNs). Unlike traditional JPEG, which is designed for human viewers, J4D aims to mi…
-
New metric predicts rPPG algorithm performance under video compression
Researchers have developed a new metric called Spatial Artifact Coherence (SAC) to predict the performance of remote photoplethysmography (rPPG) algorithms under video compression. SAC, calculated from the covariance ma…
-
SEAOTTER framework boosts robotics compression with JPEG compatibility
Researchers have developed SEAOTTER, a novel compression framework designed for cloud robotics that addresses bandwidth and compute limitations. This system combines a learned latent representation with the widely compa…
-
DocQT dataset improves document forgery detection robustness
Researchers have developed DocQT, a new dataset and method to improve the robustness of document forgery localization models. These models often fail in real-world scenarios due to a mismatch between training data and o…
-
CATRF framework improves volumetric media compression for streaming
Researchers have developed CATRF, a novel framework for compressing volumetric media to improve content delivery. This method integrates standard codecs like JPEG and AV1 directly into the training loop, allowing featur…
-
JPEG metadata can reveal AI-generated images
Researchers are exploring methods to detect AI-generated images by examining JPEG metadata, C2PA, and XMP standards. These embedded data formats can potentially reveal the origin of an image, indicating whether it was c…
-
LiVeAction neural codec offers efficient, versatile compression for real-time sensors
Researchers have developed LiVeAction, a novel neural codec designed for real-time operation on resource-constrained devices. This architecture addresses limitations of existing codecs by reducing encoder complexity thr…
-
Robotic vision uses hybrid video and stills to improve detail under bandwidth limits
Researchers have proposed a novel hybrid visual telemetry scheme for bandwidth-constrained robotic vision systems. This approach combines a continuous low-bitrate video stream with selectively transmitted high-detail st…
-
AI analyzes compressed CT scans efficiently with new FAST and SFP techniques
Researchers have developed a new framework called CT-Lite to enable AI analysis of compressed chest CT scans, addressing the computational burden of medical imaging data. The system utilizes Feature Attention Style Tran…
-
Researchers develop method to detect AI-generated content from cameras
Researchers have developed a method to detect and remove "hallucinated" content introduced by generative AI within camera image signal processors. This technique optimizes an MLP decoder and a modality-specific encoder …