InoteveraGaidhi rinotevera
Conditional GANs
AI Yekutarisa
Gwara reVisual AI
Kufambira mberi Kukura kweGANs
Kukura kunofambira mberi kunodzidzisa GAN nekutanga pane zvigadziriso zvidiki uye zvishoma nezvishoma nekuwedzera maseru kuti isvike mifananidzo ine-yepamusoro-soro.
Teerera kugwaro iriAI-yakarondedzerwa · ~2 min
Pa peji ino2 min verenga
Pfupiso
Izvo zvine basa nekuti yakaita yakagadzikana, megapixel-mhando GAN synthesis inoshanda kekutanga.
Kudzika Kwakadzika
Yakaunzwa naKarras et al. (NVIDIA) muna 2017, kukura kunofambira mberi (ProGAN) inobata kusagadzikana uye kunonoka kwekudzidzisa maGAN zvakananga pakugadziriswa kwepamusoro. Ose ari maviri jenareta nerusarura anotanga diki, pa4x4 pixels, achidzidza hombe-chimiro chimiro. Matunhu matsva anopeta gadziriso (8x8, 16x16, kusvika 1024x1024) anobva awedzerwa symmetrically kune ese ma network panguva yekudzidziswa. Zvine hutsinye, imwe neimwe nyowani nyowani inopera mushe muchishandisa mutsara alpha musanganiswa saka network haina kuvhundutswa nekukasira kwekuvaka shanduko. Nekudzidza zvakakasharara pamberi pezvakanakisa, kudzidziswa kunowedzera kugadzikana, kuchinjika nekukurumidza, uye kunogadzira zviso zvekuvimbika zvakaita kuti mhedzisiro yeCelebA-HQ ive nemukurumbira. Iro bepa rakaunzawo minibatch standard kutsauka uye yakaenzana mareti ekudzidza kuti awedzere kudzikamisa kudzidziswa.
Kuitwa muNyika yeChokwadi
Training on CelebA-HQ to generate synthetic high-resolution face samples.
Studying GAN synthesis on LSUN scene categories.
Understanding the staged architecture that the original StyleGAN extended.
Comparing coarse-to-fine schedules with later fixed-architecture GAN training.
Ramangwana Rekufambira mberi Kukura kweGANs
Kukura kunofambira mberi kwaive hwaro StyleGAN yakavakirwa pairi, asi StyleGAN2 yakazoratidza kuti chivakwa chakamisikidzwa chine skip yekubatanidza uye zvakasara zvidhinha zvaigona kuenderana nemhando yayo pasina hurongwa hwakarongwa, saka kukura kuri pachena kwakatadza kufarirwa. Nhaka yakadzama inoenderera mberi: chizvarwa chakaomarara-kune-chakanaka chave kuoneka mukupararira kwezvikero zvakawanda, mapaipi akadhindwa epamusoro-resolution, uye latent-space upscaler. Kunzwisisa kukura kunofambira mberi kunoramba kwakakosha pakubata kuti nei hierarchical, yakaderera-kusvika-yepamusoro-frequency kudzidza ichidzikamisa kudzidziswa kwekubereka.
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 Progressive Growing of GANs 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
Chii chinonzi Progressive Kukura kweGANs?
Kukura kunofambira mberi kunodzidzisa GAN nekutanga pane zvigadziriso zvidiki uye zvishoma nezvishoma nekuwedzera maseru kuti isvike mifananidzo ine-yepamusoro-soro. Izvo zvine basa nekuti yakaita yakagadzikana, megapixel-mhando GAN synthesis inoshanda kekutanga.
GAN inokura zvishoma nezvishoma inotanga sei kudzidziswa?
Kudzidzira kunotanga pa4x4, uko ma network anodzidza chimiro chakaomarara, pamberi pepamusoro-resolution layer yawedzerwa.
Ndeipi bhenefiti huru yekukura kugadzirisa zvishoma nezvishoma?
Kudzidza kwakakasharara maficha anotanga kudzikamisa kudzidziswa uye kumhanya kuchinjika zvichienzaniswa nekurwisa kuzere resolution kubva pakutanga.
Kana iyo nyowani yepamusoro-resolution layer yawedzerwa, inounzwa sei?
Linear fade-in inosanganisa iyo nyowani yekubuda neupsampled yakaderera-resolution kubuda kuitira kuti network igadzirike zvishoma nezvishoma.
Sei mitsara ichiwedzerwa kune zvese jenareta uye kusarura?
Manetiweki ese ari maviri anokura mu tandem kuitira kuti musarura arambe achiongorora mifananidzo pazvinogadziriswa jenareta.
Chii chinangwa cheiyo minibatch standard deviation layer yakaunzwa muProGAN?
Inowedzera nhamba pamusoro pekusiyana kwebatch, ichikurudzira jenareta kuti ibudise zvinobuda zvakasiyana pane kudonha.
Ramba uchidzidza
Mazano anoenderana
Mamwe madhairekitori akasarudzirwa nyaya iyi