InoteveraGaidhi rinotevera
Sora uye Zvinyorwa-ku-Vhidhiyo
AI Yekutarisa
Gwara reVisual AI
Ita-A-Vhidhiyo Chinyorwa-kune-Vhidhiyo
Make-A-Vhidhiyo ndeye Meta's 2022 sisitimu inosandura kurudziro kuita kavhidhiyo kapfupi pasina kumbodzidzira pane akanyorwa mavara-mavhidhiyo maviri.
Teerera kugwaro iriAI-yakarondedzerwa · ~2 min
Pa peji ino2 min verenga
Pfupiso
It matters because it showed that the visual knowledge inside text-to-image models could be 'taught' to move using only unlabeled video.
Kudzika Kwakadzika
Make-A-Video, announced by Meta AI in September 2022, generates brief video clips from text prompts. Its key trick is decoupling appearance from motion: a text-to-image model learns what things look like from paired text-image data, while separate spatiotemporal layers learn how they move from video footage with no associated text. Meta said this reduced the need for paired text-video examples and that the research used publicly available datasets. Meta also described making videos from images and creating new videos similar to existing videos. The base model produces low-resolution, low-frame-rate clips, then dedicated networks interpolate extra frames and upscale spatial resolution. The result was strikingly coherent for its era, though clips were short, blurry, and prone to flicker and warping.
Kuitwa muNyika yeChokwadi
Kusimudzira mutsara mumwe chete unotsanangura kuita kapfupi looping clip yesocial media post
Kuunza pfungwa yakamira senge 'teddy bear ipenda mufananidzo' kuhupenyu semufananidzo unofamba
Kupindirana pakati pemifananidzo miviri yakapihwa nemushandisi kugadzira vhidhiyo yekuchinja yakatsetseka
Kugadzira madhiramu ekukurumidza ekufamba ezviratidziro zvebhodhibhodhi pamberi pechero firimu
Ramangwana reMake-A-Vhidhiyo Chinyorwa-kune-Vhidhiyo
Make-A-Vhidhiyo's image-pre-plus-unlabeled-motion resipi yakadyara zvinyorwa-kune-vhidhiyo wave. Zvizvarwa zvaro zvinosimbisa kureba, kwepamusoro-resolution, zvimedu zvakagadzikana zvenguva pfupi zvine inodzoreka kamera inofamba uye odhiyo. Tarisira iyo yakakosha pfungwa, kushandisazve yakakura ruzivo rwemufananidzo uye kufamba kwekudzidza zvakachipa, kuenderera kunyange sezvivakwa zvinoshanduka kuenda kune transformer-based latent diffusion uye mamodheru akabatana anobvumawo mufananidzo kana vhidhiyo mamiriro ekugadzirisa nekuenderera mberi.
Kuzvishandisa
Nei zvichikosha
- Kumhanya uye chiyero. Visual AI inogona kuita otomatiki yekuongorora, yekuona, uye yekumaka mabasa pachiyero.
- Vaka sarudzo. Zvikwata zvekugadzira zvinogona prototype pfungwa nekukurumidza nekudzokororwa kwemaoko mashoma.
- Team uye workflow. Mashandisirwo anogona kushandisa masaini emifananidzo nemavhidhiyo ayo aimbove akaoma kugadzirisa.
Chenjera
- Kodzero dzemifananidzo uye kubvumirwa kunogona kuve njodzi dzepamutemo kana provenance isina kujeka.
- Kuita kwemuenzaniso kunogona kusiyanisa kupenya, huwandu hwevanhu, uye nharaunda.
- Manyepo enhema anogona kusacherechedzwa kunze kwekunge zvikumbaridzo zvekuvimba zvikatariswa.
Matanho anotevera
- Tsanangura maitiro ekugamuchirwa echokwadi, kurangarira, uye mutengo wekukanganisa.
- Edzai nedata rinoenderana nemamiriro chaiwo ekugadzira.
- Wedzera ongororo yemunhu kune yakaderera-kusavimbika kana yakakwirira-inokanganisa kufanotaura.
- Tevera modhi kudonha uye simbisa mushure mekuchinja kwekamera kana dataset.
Ramba Uchiongorora
Tsamba yemahara
Tora iyo yemazuva ese AI pfupiso
Nhatu dzakasimbiswa dzeAI nyaya mangwanani ega ega pakati pevhiki, dzakanyorwa neChirungu chakajeka. Yemahara zvachose, hapana ads.
Imwe email zuva rega rega revhiki. Bvisa kunyoresa kamwe chete. Hatimbotengese kana kugovera kero yako.
Zviedze iwe pachako
Take the Make-A-Video Text-to-Video quiz
Mhinduro yepakarepo pamhinduro yega yega, uye chitupa chinogovaniswa chine ID inobatika kana wapedza kosi.
Tsigira dzidzo yemahara yeAI. AI Understanding ndeye 501(c)(3) isingabatsiri - hapana ads, hapana paywall, zvachose. Ita chipo
Mibvunzo inowanzo bvunzwa
What is Make-A-Video Text-to-Video?
Make-A-Vhidhiyo ndeye Meta's 2022 sisitimu inosandura kurudziro kuita kavhidhiyo kapfupi pasina kumbodzidzira pane akanyorwa mavara-mavhidhiyo maviri. Izvo zvine basa nekuti zvakaratidza kuti ruzivo rwekuona mukati memavara-kune-mufananidzo modhi rinogona 'kudzidziswa' kufamba uchishandisa vhidhiyo isina kunyorwa chete.
Ndeipi yaive yepakati hunyanzvi iyo inoita kuti Make-A-Vhidhiyo irege kuda akanyorwa mavara-mavhidhiyo maviri?
Gadzira-A-Vhidhiyo inobvisa dambudziko: chinyorwa-ku-chifananidzo modhi inodzidza kuti zvinhu zvinotaridzika sei kubva pamifananidzo ine mashoko, nepo akapatsanurwa epanguva akaturikidzana achidzidza kufamba kubva kune mbishi, isina kunyorwa vhidhiyo.
Ko Make-A-Vhidhiyo inowedzera sei kugona kufamba kune mufananidzo modhi?
Iyo inowedzera 2D spatial convolutions uye kutarisisa neatsva temporal layers anodzidza machinjiro emukati nekufamba kwenguva.
Chii chinoitwa neframe-interpolation uye super-resolution matanho?
Mushure meiyo yakaderera yakaderera-res, yakaderera-frame-reti dhizaini, kududzira kunowedzera mafuremu uye super-resolution modules inosimudza kugadzirisa.
Ndechipi chaive chigadzikiso chekubuda kweMake-A-Vhidhiyo?
MaClips aingova masekonzi mashoma, akadzikira, uye aiwanzoita kupenya kwechinguva uye kukanganisa.
Besides text prompts, what visual input can Make-A-Video use to create a video?
Meta said Make-A-Video can create videos from images as well as text prompts.
Ramba uchidzidza
Mazano anoenderana
Mamwe madhairekitori akasarudzirwa nyaya iyi