Underwater object detection is critical for monitoring marine ecosystems but poses unique challenges, including degraded image quality, imbalanced class distribution, and distinct visual characteristics. Not every species is detected equally well, yet underlying causes remain unclear. We address two key research questions: 1) What factors beyond data quantity drive class-specific performance disparities? 2) How can we systematically improve detection of under-performing marine species? We manipulate the DUO and RUOD datasets to separate the object detection task into localization and classification and investigate the under-performance of the scallop class. Localization analysis using YOLO11 and TIDE finds that foreground-background discrimination is the most problematic stage regardless of data quantity. Classification experiments reveal persistent precision gaps even with balanced data, indicating intrinsic feature-based challenges beyond data scarcity and inter-class dependencies. We recommend imbalanced distributions when prioritizing precision, and balanced distributions when prioritizing recall. Improving under-performing classes should focus on algorithmic advances, especially within localization modules. We publicly release our code and datasets.

If you were around for the electropop zeitgeist of the early 2010s, chances are that Purity Ring feature prominently on your nostalgia playlist. And if you were a young adult at that time, well, there's also a high chance that you played Japanese role-playing games as a teenager – whether that was Chrono Trigger on an SNES or Final Fantasy on a PlayStation. Purity Ring's new single Many Lives is an attempt to recapture the feeling of the RPG that you discovered as a 12-year-old and immediately made into your whole personality. Inspired by games such as Skies of Arcadia, Phantasy Star Online and Secret of Mana, it is poised to tug on the heartstrings of fans of a certain vintage. This is a bold decision for a band who have previously collaborated with Deftones and covered Eurodance classics, but members Megan James and Corin Roddick have the gaming expertise to pull it off. "We're huge fans of the JRPG genre," they say, naming Nier: Automata and Final Fantasy X as major influences on the sonic atmosphere of their latest work.

Large language models (LLMs) have demonstrated the ability to improve human efficiency through conversational interactions. Conventional LLM-powered dialogue systems, operating on a turn-based paradigm, preclude real-time interaction during response generation. To address this limitation, researchers have proposed duplex models. These models can dynamically adapt to user input, facilitating real-time interactive feedback. However, these methods typically require substantial computational resources to acquire the ability. To reduce overhead, this paper presents a new duplex decoding approach that enhances LLMs with duplex ability, requiring minimal additional training. Specifically, our method employs parallel decoding of queries and responses in conversations, effectively implementing a channel-division-multiplexing decoding strategy. Experimental results indicate that our proposed method significantly enhances the naturalness and human-likeness of user-AI interactions with minimal training costs.

A few years ago I bought a Japanese copy of Snatcher, a cyberpunk video game designed by Hideo Kojima before he went on to create the legendary Metal Gear Solid series. There were just two problems with this purchase: the game is a text-heavy role-playing adventure with no English translation, and I don't own a PC Engine, a cult 16 bit machine first released in Japan in 1987, which hosted some of the finest arcade conversions of the era in then-astounding visual quality. A small number were imported into the UK, but it was never a huge hit here, so getting one on eBay is a costly and risk-laden adventure. I bought Snatcher because I loved its anime aesthetic and its role in the nascent career of a games industry legend. Last week, I loaded it up for the first ever time, thanks to the Analogue Duo.

The exponential growth of social media platforms, such as Facebook and TikTok, has revolutionized communication and content publication in human society. Users on these platforms can publish multimedia content that delivers information via the combination of text, audio, images, and video. Meanwhile, the multimedia content release facility has been increasingly exploited to propagate toxic content, such as hate speech, malicious advertisements, and pornography. To this end, content moderation software has been widely deployed on these platforms to detect and blocks toxic content. However, due to the complexity of content moderation models and the difficulty of understanding information across multiple modalities, existing content moderation software can fail to detect toxic content, which often leads to extremely negative impacts. We introduce Semantic Fusion, a general, effective methodology for validating multimedia content moderation software. Our key idea is to fuse two or more existing single-modal inputs (e.g., a textual sentence and an image) into a new input that combines the semantics of its ancestors in a novel manner and has toxic nature by construction. This fused input is then used for validating multimedia content moderation software. We realized Semantic Fusion as DUO, a practical content moderation software testing tool. In our evaluation, we employ DUO to test five commercial content moderation software and two state-of-the-art models against three kinds of toxic content. The results show that DUO achieves up to 100% error finding rate (EFR) when testing moderation software. In addition, we leverage the test cases generated by DUO to retrain the two models we explored, which largely improves model robustness while maintaining the accuracy on the original test set.

If Google's Pixel 4 is your daily driver, good news: It's now able to screen robocalls -- and more. Google announced this morning an update to the Pixel 4's Call Screen feature in the U.S. that automatically declines calls from unknown parties and filters out suspected robocallers, alongside an improved video calling experience on Duo, the rollout of the new Google Assistant to more users, and a zippier software experience made possible by memory usage optimizations. It's a part of what Google's calling feature drops, which will deliver "bigger updates" to Pixel devices with "more helpful and fun features" going forward. The first arrives starting today, with others to follow on a monthly cadence. "Pixel phones have always received monthly updates to improve performance and make your device safe," wrote Google group product manager Shenaz Zack in a blog post.

Melding the disparate worlds of art and computer science, Andreas Refsgaard and Mikkel Loose have developed a fascinating AI project called Booksby.ai, an online bookstore entirely generated by artificial intelligence. Every aspect of the site is generated by machine learning algorithms, from the entire books and accompanying cover artwork, to the reviews and pictures of people reviewing the books. And on top of that, all the books are actually available to buy on Amazon. Andreas Refsgaard is an artist exploring creative uses for machine learning. Based in Copenhagen, the Booksby.ai

The Mayo Clinic and health technology vendor Eko are working together to develop and commercialize a machine learning-based algorithm that screens patients for low ejection fraction, which is linked to heart failure. A low ejection fraction number, often measured by an echocardiogram, suggests problems with the heart's pumping function. However, echocardiography is an expensive and time-consuming medical imaging test using ultrasound that is less accessible than a doctor with a stethoscope. "With this collaboration, we hope to transform the stethoscope in the pocket of every physician and nurse from a hand tool to a power tool," said Paul Friedman, MD, chair of cardiovascular medicine at the Mayo Clinic. "The community practitioner performing high school sports physicals and the surgeon about to operate may be able to seamlessly tap the knowledge of an experienced cardiologist to determine if a weak heart pump is present simply by putting a stethoscope on a person's chest for a few seconds."

Apple has signed on It's Always Sunny in Philadelphia duo Rob McElhenney and Charlie Day for a straight-to-series comedy, according to a Variety report. The unnamed show, which is to be produced by Ubisoft, Lionsgate, and 3 Arts Entertainment, will apparently take place in a video game development studio. Ubisoft's Jason Altman, Danielle Kreinik and Gérard Guillemot (Assassin's Creed) will be credited with executive production alongside Day and McElhenney, the latter of whom is also attached to star. Hollywood Reporter says the show is a "cutting-edge" dive into the complexities of the human condition. When it comes to expanding its brigade of original TV content, Apple's not cutting any corners.

Two years ago, Ars Technica hosted the online premiere of a weird short film called Sunspring, which was mostly remarkable because its entire script was created by an AI. The film's human cast laughed at odd, computer-generated dialogue and stage direction before performing the results in particularly earnest fashion. That film's production duo, Director Oscar Sharp and AI researcher Ross Goodwin, have returned with another AI-driven experiment that, on its face, looks decidedly worse. Blurry faces, computer-generated dialogue, and awkward scene changes fill out this year's Zone Out, a film created as an entry in the Sci-Fi-London 48-Hour Challenge--meaning, just like last time, it had to be produced in 48 hours and adhere to certain specific prompts. That 48-hour limit is worth minding, because Sharp and Goodwin went one bigger this time: they let their AI system, which they call Benjamin, handle the film's entire production pipeline.