Microsoft apresenta IA que faz foto falar e se mexer

[{"selector":"#anim-52882bcd-25d0-4592-b5b6-223bfb66a745","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e29f60a4-f716-4c8a-9a5a-7baab73987a0","keyframes":{"opacity":[0,1]},"delay":0,"duration":1000,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-40bd99b0-0a4f-4ef4-b266-fe1d9e96fcb8","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-0a4d3332-63dd-47b2-8fd2-1fb3c6025ab3","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Salvatore De Lellis/Pexels

[{"selector":"#anim-61a1b22f-b8ed-4d8e-8d31-bca5442574a7","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-5f701948-388b-4886-8ede-f7c80ab6f617","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-7c47c0b5-6645-47c8-a226-5151414191ff","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-9dc535b4-05a2-430d-aa58-41774e7fff12","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Matthew Manuel/Unsplash Uma equipe de pesquisadores da Microsoft Research Lab, em Pequim, na China, apresentou o VASA-1

[{"selector":"#anim-55f38127-90fd-49f4-b149-f6e0a5394b4a","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-2a296e57-40bb-4b44-a89c-166fc87ea7ac","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-0a466e83-663c-4e6b-bd5c-a115475a83dc","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] [{"selector":"#anim-f4020127-ce39-4299-8732-aab4af34767f","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] Freepik Trata-se de um novo modelo de inteligência artificial (IA) capaz de gerar vídeos realistas em tempo real de rostos humanos a partir de uma única foto e clipe de áudio

[{"selector":"#anim-74fcbb58-e46a-4482-9b96-1ccb701c9b08","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-b3b0c5b1-1d57-42f1-9704-7885e0df8f94","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-e7dcd553-f1fe-4f28-8414-658a01e2312f","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-47812d28-d5cd-4007-8669-2d3f2cf22678","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Divulgação O resultado são vídeos curtos que dão à imagem estática movimentos, expressões faciais e sincronia labial extremamente convincentes, capazes de enganar um espectador menos atento

[{"selector":"#anim-849d0693-656f-437e-868c-efc6fa1f96b8","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-9788a66a-3d26-4ccf-a29b-45a59d582d32","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-18365cf8-173a-41bb-bc54-bd029b92ebf0","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-6432129c-0439-4780-920f-1e147a320a18","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Divulgação A tecnologia também é capaz de animar imagens artísticas, como personagens fictícios e obras de arte, além de gerar falas e cantorias

[{"selector":"#anim-bba54f9e-8fea-4104-8205-77fddcbbbbc1","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-34cb5f99-dd61-4833-b76d-143c8ad7bbf7","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-867b1431-f37c-45cd-8029-2a1ea0b2f405","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-de6f9cde-aa23-4f66-8958-b3ed1f9d1395","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Para a geração de fala, é preciso incluir um clipe de áudio da voz desejada (que pode ser do indivíduo da foto ou não) Freepik

[{"selector":"#anim-571798f4-0182-4554-8e44-cd7124b92fdd","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-9036d9e5-16dc-41f9-9bf6-9682ffbee293","keyframes":{"opacity":[0,1]},"delay":0,"duration":1500,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-30444b17-4627-40bf-837e-1213fdfb82bd","keyframes":{"opacity":[0,1]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"both"}] [{"selector":"#anim-0ae18245-d341-4bdc-86ae-92f9148ae52a","keyframes":{"transform":["scale(0.3333333333333333)","scale(1)"]},"delay":0,"duration":500,"easing":"cubic-bezier(.3,0,.55,1)","fill":"forwards"}] Divulgação A ferramenta também dá ao usuário controle de expressões faciais, direção do olhar, distância da cabeça, ângulo da câmera e outros ajustes granulares

CNNBRASIL.COM.BR

[{"selector":"#anim-14d3921a-125d-4219-bf21-d2a082666efc","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a6a78b3c-dbe8-4c9d-890a-688042f8a893","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.4, 0.4, 0.0, 1)","fill":"both"}] [{"selector":"#anim-a9579a77-80d0-4f9f-bf83-c7fc972d7da1","keyframes":{"opacity":[0,1]},"delay":0,"duration":1000,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-718830b4-1e8e-44b0-a590-247d4672a081","keyframes":{"transform":["translate3d(0px, 2132.63615%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-f24c4bfd-0f98-4d8c-88f2-21344f250d50","keyframes":{"opacity":[0,1]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-9c98ca59-3b13-4804-9c02-035c7b7eb6b3","keyframes":{"transform":["translate3d(-117.84387%, 0px, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":600,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-0524a89a-8e52-48c8-831a-3e11bcab6850","keyframes":{"opacity":[0,1]},"delay":0,"duration":1000,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] [{"selector":"#anim-fc872c46-57e4-41de-abe9-c136b3290914","keyframes":{"transform":["translate3d(0px, 1223.93157%, 0)","translate3d(0px, 0px, 0)"]},"delay":0,"duration":1000,"easing":"cubic-bezier(0.2, 0.6, 0.0, 1)","fill":"both"}] leia mais em Divulgação