Speech foundation models (SFMs) have demonstrated strong performance across a variety of downstream tasks, including speech intelligibility prediction for hearing-impaired people (SIP-HI). However, optimizing SFMs for SIP-HI has been insufficiently explored. In this paper, we conduct a comprehensive study to identify key design factors affecting SIP-HI performance with 5 SFMs, focusing on encoder layer selection, prediction head architecture, and ensemble configurations. Our findings show that, contrary to traditional use-all-layers methods, selecting a single encoder layer yields better results. Additionally, temporal modeling is crucial for effective prediction heads. We also demonstrate that ensembling multiple SFMs improves performance, with stronger individual models providing greater benefit. Finally, we explore the relationship between key SFM attributes and their impact on SIP-HI performance. Our study offers practical insights into effectively adapting SFMs for speech intelligibility prediction for hearing-impaired populations.

Like most other tech companies, Google is investing heavily in the development of AI models and trying to incorporate AI into anything and everything in their portfolio. The latest endeavor apparently involves Google integrating its Gemini AI assistant into its world-popular Chrome browser--at least, that's the rumor going around. And that rumor is backed up by some newly discovered code in the latest version of Chrome Canary, reports Windows Latest. Canary is a special version of the browser for testing out experimental features, and it appears that Gemini is being integrated with it. However, it doesn't seem to be fully operational yet. The new feature is called GLIC, which stands for "Gemini Live in Chrome," and it comes with a new "Glic" section in Chrome's settings page.

How much information about training samples can be gleaned from synthetic data generated by Large Language Models (LLMs)? Overlooking the subtleties of information flow in synthetic data generation pipelines can lead to a false sense of privacy. In this paper, we design membership inference attacks (MIAs) that target data used to fine-tune pre-trained LLMs that are then used to synthesize data, particularly when the adversary does not have access to the fine-tuned model but only to the synthetic data. We show that such data-based MIAs do significantly better than a random guess, meaning that synthetic data leaks information about the training data. Further, we find that canaries crafted to maximize vulnerability to model-based MIAs are sub-optimal for privacy auditing when only synthetic data is released. Such out-of-distribution canaries have limited influence on the model's output when prompted to generate useful, in-distribution synthetic data, which drastically reduces their vulnerability. To tackle this problem, we leverage the mechanics of auto-regressive models to design canaries with an in-distribution prefix and a high-perplexity suffix that leave detectable traces in synthetic data. This enhances the power of data-based MIAs and provides a better assessment of the privacy risks of releasing synthetic data generated by LLMs.

It was observed in [6] that such long utterances harm the model convergence. We also note that this Recent advances in speech recognition and translation rely on approach may lead to significant padding in mini-batches, resulting hundreds of thousands of hours of Internet speech data. We argue in wasted computation on non-informative frames. We that state-of-the art accuracy can be reached without relying on present an alternative approach to sampling and batching that web-scale data. Canary - multilingual ASR and speech translation allows us to iterate through data twice as fast, while balancing model, outperforms current state-of-the-art models - Whisper, different languages and data sources better. We further accelerate OWSM, and Seamless-M4T on English, French, Spanish, and the training and inference by adopting a FastConformer [7] architecture German languages, while being trained on an order of magnitude and initializing the encoder from a ASR only pretrained less data than these models. Three key factors enables such dataefficient checkpoint.

When users open an Incognito browser on Chrome, they'll see a notification warning them that other people using their device won't be able to see their activity but that their downloads, bookmarks and reading items will still be saved. Now, Google has updated that disclaimer in Chrome's experimental Canary channel, shortly after agreeing to settle a 5 billion lawsuit accusing it of tracking Incognito users. As first noticed by MSPowerUser, the company has tweaked the disclaimer in Canary to add language that says Incognito mode won't change how websites collect people's data. "Others who use this device won't see your activity, so you can browse more privately," the new disclaimer reads. "This won't change how data is collected by websites you visit and the services they use, including Google. Downloads, bookmarks and reading list items will be saved."

Most existing dialogue systems fail to respond properly to potentially unsafe user utterances by either ignoring or passively agreeing with them. To address this issue, we introduce ProsocialDialog, the first large-scale multi-turn dialogue dataset to teach conversational agents to respond to problematic content following social norms. Covering diverse unethical, problematic, biased, and toxic situations, ProsocialDialog contains responses that encourage prosocial behavior, grounded in commonsense social rules (i.e., rules-of-thumb, RoTs). Created via a human-AI collaborative framework, ProsocialDialog consists of 58K dialogues, with 331K utterances, 160K unique RoTs, and 497K dialogue safety labels accompanied by free-form rationales. With this dataset, we introduce a dialogue safety detection module, Canary, capable of generating RoTs given conversational context, and a socially-informed dialogue agent, Prost. Empirical results show that Prost generates more socially acceptable dialogues compared to other state-of-the-art language and dialogue models in both in-domain and out-of-domain settings. Additionally, Canary effectively guides conversational agents and off-the-shelf language models to generate significantly more prosocial responses. Our work highlights the promise and importance of creating and steering conversational AI to be socially responsible.

Online fraud and identity theft is a global problem that has only been exacerbated with increased online transactions amid the COVID-19 pandemic. In particular, it is estimated that Brazilian companies lose over $41 billion due to fraud every year. In an attempt to tackle this problem head on, Lincoln Ando and Raphael Melo started idwall in mid-2016. São Paulo-based idwall started as an automated background check solution and has since grown into a suite of data and identity validation and risk analysis products. For the consumer market, its "MeuID" app is aimed at users who want to change the way they identify themselves and share their data.

While Canary's security cameras can notify you when they detect something moving in your home, they can't differentiate between an intruder and your pet Fluffy doing zoomies. Once their new feature rolls out, though, you won't have to get 20 notifications in an hour if you don't want to. The company has announced that it's rolling out Person Detection to all Canary and Canary Flex cameras in the near future -- for free. It relies on machine learning to figure out whether your camera is seeing a human being, so the system can send you specific person alerts. Canary didn't say when the feature will be available.

It's been a bit of a bumpy road for Canary and its connected security cameras. While Rita had mostly nice things to say about the all-in-one camera in her review, the company has since done its best to annoy users (particularly those on the free tier) by changing its subscription model and then backtracking slightly. Perhaps some new features will go some way to win back the trust of its users. 'Canary Vision' is a new suite of advanced detection capabilities that will be made available to both free and paid subscribers in the coming weeks. Powered by AI and machine learning, intelligent person detection will be coming to Canary's all-in-one and Flex camera systems.