人工智能

489 bookmarks
Custom sorting
Semantic reconstruction of continuous language from non-invasive brain recordings | bioRxiv
Semantic reconstruction of continuous language from non-invasive brain recordings | bioRxiv
A brain-computer interface that decodes continuous language from non-invasive recordings would have many scientific and practical applications. Currently, however, decoders that reconstruct continuous language use invasive recordings from surgically implanted electrodes[1][1]–[3][2], while decoders that use non-invasive recordings can only identify stimuli from among a small set of letters, words, or phrases[4][3]–[7][4]. Here we introduce a non-invasive decoder that reconstructs continuous natural language from cortical representations of semantic meaning[8][5] recorded using functional magnetic resonance imaging (fMRI). Given novel brain recordings, this decoder generates intelligible word sequences that recover the meaning of perceived speech, imagined speech, and even silent videos, demonstrating that a single language decoder can be applied to a range of semantic tasks. To study how language is represented across the brain, we tested the decoder on different cortical networks, and found that natural language can be separately decoded from multiple cortical networks in each hemisphere. As brain-computer interfaces should respect mental privacy[9][6], we tested whether successful decoding requires subject cooperation, and found that subject cooperation is required both to train and to apply the decoder. Our study demonstrates that continuous language can be decoded from non-invasive brain recordings, enabling future multipurpose brain-computer interfaces. [1]: #ref-1 [2]: #ref-3 [3]: #ref-4 [4]: #ref-7 [5]: #ref-8 [6]: #ref-9
·biorxiv.org·
Semantic reconstruction of continuous language from non-invasive brain recordings | bioRxiv
Introducing Whisper
Introducing Whisper
We’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition.
·openai.com·
Introducing Whisper
DreamFusion: Text-to-3D using 2D Diffusion
DreamFusion: Text-to-3D using 2D Diffusion
We combine neural rendering with a multi-modal text-to-2D image diffusion generative model to synthesize diverse 3D objects from text.
·dreamfusion3d.github.io·
DreamFusion: Text-to-3D using 2D Diffusion
Lens Protocol
Lens Protocol
Lens Protocol is a composable and decentralized social graph, ready for you to build on so you can focus on creating a great experience, not scaling your users.
·lens.xyz·
Lens Protocol
Decibel
Decibel
Decibel is an early-stage venture capital firm. We invest in essential infrastructure software used by developers, data engineers, and cybersecurity teams.
·decibel.vc·
Decibel
Replicate
Replicate
Run open-source machine learning models with a cloud API
·replicate.com·
Replicate
young-geng/EasyLM: Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.
young-geng/EasyLM: Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.
Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax. - GitHub - young-geng/EasyLM: Large language models (LLM...
·github.com·
young-geng/EasyLM: Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.
Sudowrite
Sudowrite
Bust writer's block and be more creative with our magical writing AI.
·sudowrite.com·
Sudowrite
Flowrite - Supercharge your daily communication
Flowrite - Supercharge your daily communication
Flowrite turns instruction into ready-to-send emails and messages across the browser. Save time · Hit the right tone · Overcome the blank page syndrome · Sound fluent in English.
·flowrite.com·
Flowrite - Supercharge your daily communication