Recent advancement in multimodal LLMs (MLLMs) has demonstrated their remarkable capability to generate descriptive captions for input videos. However, these models suffer from factual inaccuracies in the generated descriptions, causing severe hallucination issues. While prior works have explored alleviating hallucinations for static images, jointly mitigating visual object and temporal action hallucinations for dynamic videos remains a challenging and unsolved task. T o tackle this challenge, we propose a Self-Augmented Contrastive Alignment (SANTA) framework for enabling object and action faithfulness by exempting the spurious correlations and enforcing the emphasis on visual facts. SANTA employs a hallucinative self-augmentation scheme to identify the potential hallucinations that lie in the MLLM and transform the original captions to the contrasted negatives. Furthermore, we develop a tracklet-phrase contrastive alignment to match the regional objects and relation-guided actions with their corresponding visual and temporal phrases. Extensive experiments demonstrate that SANTA outperforms existing methods in alleviating object and action hallucinations, yielding superior performance on the hallucination examination benchmarks.

AI slop is flooding every single digital platform, and music streaming services are no exception--so much so, even someone who generally avoids AI might find themselves unknowingly listening to a robot hornily singing about butts. Take the sordid saga of "Make Love to My Shitter," an AI-generated track from an artist called BannedVinylCollection. Brace Belden, a host of the popular politics podcast TrueAnon, says that Spotify recently queued up the bawdy song after he'd finished listening to alt-country legend Lucinda Williams' 1992 album Sweet Old World. "I didn't realize the song was AI at first," he says. "I thought it might've been some obscene joke record from the 80s or 90s."

Every year since 1955, the North American Air Defense Command (NORAD) simulates the path of Santa Claus as he leaves the North Pole and delivers gifts to children around the world. For Christmas 2024, NORAD has added an AI chatbot to its tracking site, although its AI capabilities are quite limited in scope, according to an initial test. NORAD Tracks Santa officially started on December 1, but the actual simulation starts at midnight on December 24. Check out the NORAD Tracks Santa site in advance to play web games, watch videos about Santa and NORAD, listen to Santa's favorite holiday songs, learn more about Santa and his traditions, discover what NORAD does, and chat with Radar (the AI chatbot). The real-time simulation can take on martial proportions, as this video from Christmas 2013 shows.

Scientists have created a Santa-detection machine and used it to prove what children have been telling adults for generations – that Santa has a unique face which clearly distinguishes him from other elderly bearded men. Previous research has suggested that children as young as three can identify Santa Claus based on his distinctive appearance. "Santa Claus or Father Christmas is obviously a very recognisable person, but we wondered whether this is because of his clothing, his beard or his face?" said Dr Thomas Wright, a clinical geneticist at the University of Manchester, who led the research. "Until now no one has ever systematically and scientifically interrogated the modern face of Father Christmas." To investigate, Wright and his colleagues trained a deep-learning facial-recognition algorithm – similar to the facial ID software used to unlock smartphones or identify people in digital photo libraries – on multiple images of Santa gathered from the internet.

From Richard Attenborough in'Miracle on 34th Street' to Kurt Russell in'The Christmas Chronicles' a number of famous actors have taken on the role of Santa Claus in blockbuster hits through the years. But who would take on the leading role if Hollywood cast a new movie featuring Father Christmas? To answer this burning question, MailOnline turned to ChatGPT. While the AI bot says that casting for a dream Santa would depend on the tone and style of the film, it suggests five actors who could take on the role. So, do you agree with its star-studded suggestions?

Some people are definitely ending up on the naughty list this Christmas. Erotic audio platform BLOOM now lets users'sext' with Father Christmas himself. To celebrate the festive season, users of the bizarre AI chat service will now be able to speak with a'hot Santa' or his'naughty elf'. The AI will not only send flirty texts but even create unique audio messages in the voice of the Christmas characters. So if you're feeling frisky this festive season, you can now live out your fantasy from the comfort of your own phone.

Every single year I tell myself that I'm not going to go overboard for the holidays. Then I start shopping for all 20 of my 20 closest family members and friends. Pair that big list of people with the fact that gift-giving is my love language, and you've got a recipe for blowing through a budget. But this year, I mean it! If you're in the same boat, you'll be happy to know that there are plenty of reviewer-approved gadgets and goodies available for $50 or less.

Distantly-Supervised Named Entity Recognition effectively alleviates the burden of time-consuming and expensive annotation in the supervised setting. But the context-free matching process and the limited coverage of knowledge bases introduce inaccurate and incomplete annotation noise respectively. Previous studies either considered only incomplete annotation noise or indiscriminately handle two types of noise with the same strategy. In this paper, we argue that the different causes of two types of noise bring up the requirement of different strategies in model architecture. Therefore, we propose the SANTA to handle these two types of noise separately with (1) Memory-smoothed Focal Loss and Entity-aware KNN to relieve the entity ambiguity problem caused by inaccurate annotation, and (2) Boundary Mixup to alleviate decision boundary shifting problem caused by incomplete annotation and a noise-tolerant loss to improve the robustness. Benefiting from our separate tailored strategies, we confirm in the experiment that the two types of noise are well mitigated. SANTA also achieves a new state-of-the-art on five public datasets.

Facial recognition for sleigh ignition is unable to recognize Santa consistently through beard and hat. Range is a little less than four hundred miles. Nearest village to the North Pole is about eight hundred miles away. "Frunk" full of presents keeps catching fire. Reindeer have begun to protest what they believe to be unlawful termination.

Sign up to receive the Future Tense newsletter every other Saturday. It's a busy time to run a Secret Santa site. For Elfster, one such site, the normal staff of around 20 elves (yes, they call themselves elves) balloons to a team of about 55. "We need a million servers, we need tons of support people--it's just off the hook for the holidays," Peter Imburg, the company's CEO, told me recently on Zoom. More than 21 million people have used Elfster for exchanges, and Imburg said the biggest group he's seen participate in an exchange on the platform was about 5,000 people. Imburg started Elfster in 2004 (the time of Friendster and Napster, hence the name) as a way to solve a personal problem: How to coordinate a family gift exchange when not everyone was in the same place, and with certain conditions, like not selecting yourself or your spouse?