Search Results

< Back SeamlessM4T Massively Multilingual And Multimodal Machine Translation While recent breakthroughs in text-based models have pushed machine translation coverage beyond 200 languages, unified speech-to-speech translation models have yet to achieve similar strides. Meta Facebook 22 Aug 2023 What does it take to create the Babel Fish, a tool that can help individuals translate speech between any two languages? While recent breakthroughs in text-based models have pushed machine translation coverage beyond 200 languages, unified speech-to-speech translation models have yet to achieve similar strides. More specifically, conventional speech-to-speech translation systems rely on cascaded systems composed of multiple subsystems performing translation progressively, putting scalable and high-performing unified speech translation systems out of reach. To address these gaps, we introduce SeamlessM4T—Massively Multilingual & Multimodal Machine Translation—a single model that supports speech-to-speech translation, speech-to-text translation, text-to-speech translation, text-to-text translation, and automatic speech recognition for up to 100 languages. To build this, we used 1 million hours of open speech audio data to learn self-supervised speech representations with w2v-BERT 2.0. Subsequently, we created a multimodal corpus of automatically aligned speech translations, dubbed SeamlessAlign. Filtered and combined with human labeled and pseudo-labeled data (totaling 406,000 hours), we developed the first multilingual system capable of translating from and into English for both speech and text. On Fleurs, SeamlessM4T sets a new standard for translations into multiple target languages, achieving an improvement of 20% BLEU over the previous state-of-the-art in direct speech-to-text translation. Compared to strong cascaded models, SeamlessM4T improves the quality of into-English translation by 1.3 BLEU points in speech-to-text and by 2.6 ASR-BLEU points in speech-to-speech. On CVSS and compared to a 2-stage cascaded model for speech-to-speech translation, SeamlessM4T-Large’s performance is stronger by 58%. Preliminary human evaluations of speech-to-text translation outputs evinced similarly impressive results; for translations from English, XSTS scores for 24 evaluated languages are consistently above 4 (out of 5). For into English directions, we see significant improvement over WhisperLarge-v2’s baseline for 7 out of 24 languages. To further evaluate our system, we developed Blaser 2.0, which enables evaluation across speech and text with similar accuracy compared to its predecessor when it comes to quality estimation. Tested for robustness, our system performs better against background noises and speaker variations in speech-to-text tasks (average improvements of 38% and 49%, respectively) compared to the current state-of-the-art model. Critically, we evaluated SeamlessM4T on gender bias and added toxicity to assess translation safety. Compared to the state-of-the-art, we report up to 63% of reduction in added toxicity in our translation outputs. Finally, all contributions in this work—including models, inference code, finetuning recipes backed by our improved modeling toolkit Fairseq2, and metadata to recreate the unfiltered 470,000 hours of SeamlessAlign — are open-sourced and accessible at https://github.com/facebookresearch/seamless_communication . https://ai.meta.com/research/publications/seamlessm4t-massively-multilingual-multimodal-machine-translation/ For more: NLP, AI, Machine Learning, Audio, Standards Previous Next

All Improv Perfomers All Improv Gigs Ronnie McGrath Poetry Performer Bio (Max 1000 characters) London and South East improv gigs Boat-Ting Oct18 1 October 2018 Boat-Ting Oct 2018 http://Website or social media email or phone number

< Back A Bold New Plan For Preserving Online Privacy And Security Decoupling our identities from our data and actions could safeguard our secrets IEEE Spectrum 27 Dec 2023 Three classes of new technology developed in the last few years now make decoupling practical in many more applications. Imagine you’re on a Zoom call. Your device and those of your colleagues are sending video to Zoom’s servers. By default, this is encrypted when sent to Zoom, but Zoom can decrypt it. That means Zoom’s servers see the video and hear the audio, and then forward it to others on the call. Zoom also knows who’s talking to whom, and when. Meetings that were once held in a private conference room are now happening in the cloud, and third parties like Zoom see it all: who, what, when, where. There’s no reason a videoconferencing company has to learn such sensitive information about every organization it provides services to. But that’s the way it works today, and we’ve all become used to it. There are multiple threats to the security of that Zoom call. A Zoom employee could go rogue and snoop on calls. Zoom could spy on calls of other companies or harvest and sell user data to data brokers. It could use your personal data to train its AI models. And even if Zoom and all its employees are completely trustworthy, the risk of Zoom getting breached is omnipresent. Whatever Zoom can do with your data in motion, a hacker can do to that same data in a breach. Decoupling data in motion could address those threats. Videoconferencing doesn’t need access to unencrypted video to push bits between your device and others. A properly decoupled video service could secure the who, what, where, and when of your data in motion, beginning with the “what”—the raw content of the call. True end-to-end encryption of video and audio would keep that content private to authorized participants in a call and nobody else. (Zoom does currently offer this option, but using it disables many other features.) https://spectrum.ieee.org/data-privacy For more: AI, Machine Learning, Cloud, Security, Protocols, GPU, TEE, CPU, Decoupling, Algorithm, Design Previous Next

All Improv Perfomers All Improv Gigs Phil Durrant Electric Mandolin Performer Bio (Max 1000 characters) London and South East improv gigs Boat-Ting Dec22 5 December 2022 Boat-Ting Dec 2022 http://Website or social media email or phone number

Gravesend Innovation Showcase 2024 events, locations, talks, speakers and times. < Back Nodevember Start Time 1 November Running Time 1 Month Location Across Town Centre Event Type Participation About the Event Introduction Nodevember is a global challenge to learn, share, and connect through procedural programming with tools like Blender. Running throughout November as part of the Innovation Showcase. The prompts were released on Friday, 01 November 2024 The first two events will be at the Fourth Portal, Royal Pier Road, Gravesend, DA12 2BD Friday, 1st November from 6:30 until we are bored or cold Sunday 3rd November from 3pm until we are bored or cold :) We will then decide other potential locations... " More on the event lead The event will be led, at the beginning at least, by local Gravesend programmer Oli Donald. He intends to respond to the prompts but this year in a public space and invites anyone else, even vaguely interested to join him. Event Information These events are free and come and go as you please. We intend to host this every other day in different locations around Gravesend. You may need a computer of some type if you wish to participate, but where possible Fourth Portal will have some Macs available. No understanding of coding is required. You are welcome to join in, ask questions or just watch. There is no formal training taking place, this is just a fun and enjoyable event - although you will probably learn loads! We will not know the prompts either until the 1st November and then will seek to find matching locations - might be fun! ___ Even more information ___ The official prompt list was released on the 1st November 2024. Nodevember is all about procedural awesomeness! Nodevember is a yearly challenge during the month of November, to improve and show off your procedural skills. Each year we release a list of prompts to inspire your node-based creations throughout the month. Anything procedural is welcome. If you can call it nodes, you can call it Nodevember!" The prompt list from 2023 is below. To learn more see https://nodevember.io/#participate This event is being updated Previous Next

All Improv Perfomers All Improv Gigs Jess Bull Anderson Trombone Performer Bio (Max 1000 characters) London and South East improv gigs London Improvisers Orchestra #39 28 September 2025 Performer notes, message or style description http://Website or social media email or phone number

All Improv Perfomers All Improv Gigs Myco Phoonk Words, Music & Myco Sounds Performer Bio (Max 1000 characters) London and South East improv gigs ImproVox #03 30 September 2024 Performer notes, message or style description http://Website or social media email or phone number

Gravesend Innovation Showcase 2024 events, locations, talks, speakers and times. < Back Meet Mechanismo - Engaging and Playful Robots Start Time 11:00 Running Time All Day Location 1-2, 14, 17 Popping Up Event Type Demonstration About the Event Introduction The Mechanismo team will be demonstrating their troupe of joyful robots that incorporate advanced sensing technology, four-wheel drive, and expressive movements. Come take one of the robots for a remote-controlled test drive, learn about what it takes to build a bot, and discover the surprises that emerge when humans and robots play together. More on the event lead Building on a lifetime of electronics engineering and digital fabrication, Andrew Payne founded Mechanismo in 2017 and has acted as Director ever since. He previously founded HexRay, a company specialis ing in CAD/CAM software with thousands of users worldwide. He is an accomplished artist and engineer, often exploring the themes of sustainability, biology and optimistic futures in his work. Nicky Cure is Mechanismo's Artistic Director. As an experienced creative practitioner with a background in arts education, Nicky employs "process-hacking" as a way to develop new, sustainable, and interdisciplinary methods and projects. She is an advocate for playful learning, and outside her work with Mechanismo, she is a musician and filmmaker. Event Information This event is Free All ages This event is ideal for tech enthusiasts, robotics hobbyists, families, educators, and anyone interested in playful, hands-on experiences with robots. It appeals to those curious about robotics, digital fabrication, and creative engineering, as well as those keen to see the intersection of art and technology in a fun, interactive setting. ___ Even more information ___ See the Crawly Bots dashing around the Fourth Portal in January 2024 Previous Next

All Improv Perfomers All Improv Gigs Sid Phipps Performer Bio (Max 1000 characters) London and South East improv gigs SKRONK #178 16 September 2025 Performer notes, message or style description http://Website or social media email or phone number

< Back New AI Beats DeepMind’s AlphaGo Variants 97% Of The Time! WATCH: Adversarial attack paper showing how to confuse a competent AI and re-programme the AI brain. YouTube 21 Aug 2023 Abstract We attack the state-of-the-art Go-playing AI system KataGo by training adversarial policies against it, achieving a >97% win rate against KataGo running at superhuman settings. Our adversaries do not win by playing Go well. Instead, they trick KataGo into making serious blunders. Our attack transfers zero-shot to other superhuman Go-playing AIs, and is comprehensible to the extent that human experts can implement it without algorithmic assistance to consistently beat superhuman AIs. The core vulnerability uncovered by our attack persists even in KataGo agents adversarially trained to defend against our attack. Our results demonstrate that even superhuman AI systems may harbor surprising failure modes. The papers in this episode are available here: https://goattack.far.ai/ https://arxiv.org/abs/2211.00241 YouTube: Two Minute Papers . https://www.youtube.com/watch?v=UTUnVEiO-TU For more: AI, Machine Learning, Algorithm, LLM Previous Next

< Back Linked Data Search For Donors To Leopold McClintock's 1857 Franklin search expedition Wikidata, linked data, indirectly published. The Fourth Portal Lab hosts a plaque dedicated to the crew lost searching for the Northwest Passage in 1845. This PhD research uses a form of linking data to decipher the donors who paid for a search party to go out and find the ships. All Well Polar Notes by Allegra Rosenberg 7 Dec 2022 Example of linking data The Fourth Portal Lab hosts a plaque dedicated to the crew lost searching for the Northwest Passage in 1845. This PhD research uses a form of linking data to decipher the donors who paid for a search party to find the ships. ___ Using the list included in the book The Voyage of the Fox, I sought out biographical detail for as many of the donors to Leopold McClintock’s 1857 expedition in search of Sir John Franklin as I could find. The list is a fascinating cross-section of Victorian society: from artists to Chartists, musicians to engineers, Navy and Army men to Bishops, all invested in the ongoing public project that was the search for Franklin. There are many rich peers and notable MPs on the list, but there are also many nigh-anonymous Esqs. and Misses who gave just a few shillings. Naturally I was able to find far more biographical detail on the wealthier and more famous donors—however, in opening this up as a Public Humanities project I hope to receive comments and input that will expand the available information to hopefully include more details about the unidentified donors. ___ Full details here The Donors of the Fox Posted on December 7, 2022 by admin This summer, I was the recipient of the generous Digital Humanities Fellowship from NYU. My proposed project was a broad-reaching initiative called “Visualizing the Victorian Polar Network,” but once the fellowship began I quickly realized I had to narrow my view, and ended up settling on a revised version of my project where I would create an interactive, annotated biographical database of all of the donors to Leopold McClintock’s 1857 Franklin search expedition, spearheaded by Lady Jane Franklin as a private expedition to look for her missing husband and his crew after the Admiralty had abdicated all responsibility for the continuing search. cont ... https://allegrarosenberg.notion.site/9d3c4db5a6f743668405ef7166729358?v=25c9bbaa820c4542b8e1a6278f5ec7c7 For more: Machine Learning, Linked Data, Meta Tags Previous Next

All Improv Perfomers All Improv Gigs Kay Grant Voice Performer Bio (Max 1000 characters) London and South East improv gigs Flimflam 28 May 2025 Boat-Ting Feb19 4 February 2019 Boat-Ting Sep24 2 September 2024 ImproVox #03 30 September 2024 Performer notes, message or style description http://Website or social media email or phone number