Large vision-language models (LVLMs) have significantly improved multimodal reasoning tasks, such as visual question answering and image captioning. These models embed multimodal facts within their parameters, rather than relying on external knowledge bases to store factual information explicitly. However, the content discerned by LVLMs may deviate from actual facts due to inherent bias or incorrect inference. To address this issue, we introduce MFC-Bench, a rigorous and comprehensive benchmark designed to evaluate the factual accuracy of LVLMs across three tasks: Manipulation, Out-of-Context, and Veracity Classification. Through our evaluation on MFC-Bench, we benchmarked 12 diverse and representative LVLMs, uncovering that current models still fall short in multimodal fact-checking and demonstrate insensitivity to various forms of manipulated content. We hope that MFC-Bench could raise attention to the trustworthy artificial intelligence potentially assisted by LVLMs in the future. The MFC-Bench and accompanying resources are publicly accessible at https://github.com/wskbest/MFC-Bench, contributing to ongoing research in the multimodal fact-checking field.

Scene categorization is a useful precursor task that provides prior knowledge for many advanced computer vision tasks with a broad range of applications in content-based image indexing and retrieval systems. Despite the success of data driven approaches in the field of computer vision such as object detection, semantic segmentation, etc., their application in learning high-level features for scene recognition has not achieved the same level of success. We propose to generate a fast and efficient intermediate interpretable generalized global descriptor that captures coarse features from the image and use a classification head to map the descriptors to 3 scene categories: Rural, Urban and Suburban. We train a Variational Autoencoder in an unsupervised manner and map images to a constrained multi-dimensional latent space and use the latent vectors as compact embeddings that serve as global descriptors for images. The experimental results evidence that the VAE latent vectors capture coarse information from the image, supporting their usage as global descriptors. The proposed global descriptor is very compact with an embedding length of 128, significantly faster to compute, and is robust to seasonal and illuminational changes, while capturing sufficient scene information required for scene categorization.

If you're just starting out in Intelligent Automation (IA) or Robotic Process Automation (RPA), it won't be long before you start hearing a certain acronym banded around again and again and again. Indeed, the RPA Centre of Excellence (CoE) retains a special importance in the IA/RPA universe. But what exactly is a CoE? Why should you have one? And, more to the point, how do you go about setting one up? Edge Tech's US Consultant, Greg Hunt, sat down for a chat with Andy Fanning, an RPA leader/executive based out of Lake Ozark, Missouri to dig deep into this topic. Before we dive in, it'd be great for you to set the scene. If you think back to the second industrial revolution when electricity was replacing steam power in the factories, at what point could it have been said that the revolution started? When was the tipping point?

Netflix is raising its subscription prices for all 58 million of its US users, marking the first increase since 2017. It marks the biggest price increase for monthly subscriptions since launching its streaming service in 2007, and the first since 2017. All new Netflix subscribers will be subject to the increased prices, while existing subscribers will see their subscriptions go up over the next three months. Netflix has made a major effort in recent years to produce original content in order to keep subscribers loyal to its platform and not switch to rivals like Amazon Prime. A cartoon about a talking horse, starring the goofy older brother from Arrested Development… on paper little about BoJack Horseman screams "must watch". Yet the series almost immediately transcended its format to deliver a moving and very funny rumination on depression and middle-age malaise. Will Arnett plays BoJack – one time star of Nineties hit sitcom Horsin' Around – as a lost soul whose turbo-charged narcissism prevents him getting his life together. Almost as good are a support cast including Alison Brie (Glow, Mad Men), Aaron Paul, of Breaking Bad, and Amy Sedaris as a pampered Persian cat who is also BoJack's agent.