Mip-NeRF iyo Goobaha Shucaaca ee Ka-hortagga Aliased
Mip-NeRF waxay hagaajisaa muuqalka blushka ah, ee jeexjeexan ee aafeeya NeRF-kii asalka ahaa marka aad soo bandhigto muuqaallo masaafo kala duwan ama xalal kala duwan ah.
Dulmar
It does this by tracing cones instead of infinitely thin rays, making 3D scene rendering both sharper and faster to train.
quusid qoto dheer
NeRF-kii asalka ahaa wuxuu sawiraa muuqaal ay la socdaan fallaadho dhuuban, hal dhibic markiiba, wuxuuna quudiyaa boos kasta oo 3D ah shabakad neerfaha ah. Dhibaatada: hal dhibic ayaa iska indhatiraysa inta goobta uu pixel-ku dhab ahaantii daboolayo. pixel meel u dhow kamarada waxay arkaysaa gobol yar; isla pixel ka fog ayaa arkaya mid weyn. Samaynta iyaga si isku mid ah waxay sababtaa kala-soocidda - liidashada iyo jagooyinka markaad soo dhawaajiso ama dhaqaaqdo. Mip-NeRF (Barron et al., 2021) waxay ku beddeshaa raajo kasta kolay waxayna u kala qaybisaa jahawareerka conical. Halkii laga codayn lahaa dhibic, waxay ku dhejinaysaa gobolka gudaha jahwareer kasta iyadoo la adeegsanayo codayn meelaysan oo isku dhafan (IPE), iyadoo ku qiyaaseysa mugga Gaussian. Tani waxay u oggolaanaysaa hal shabakad oo miisaan badan inay u keento xallin kasta si nadiif ah, gooynta qaladka iyo waqtiga tababarka si weyn.
Aragtida Farsamada
Khiyaamada furuhu waa codayn meelaysan oo isku dhafan. Heerka caadiga ah ee NeRF waxa uu dhibco ku sameeyaa sine iyo cosine-ka hawlo badan oo isdaba joog ah. Mip-NeRF beddelkeeda waxa ay ku qiyaastaa niyad-jabka conical sida Gaussian-gaab badan oo kala duwan waxayna xisaabisaa qiimaha la filayo ee sinusoids-kaas Gaussian. Tilmaamaha soo noqnoqda ee aad u kala duwan oo ku kala duwan gudaha jahwareerka weyn ayaa si toos ah loo dhimay eber, ilaa hadda ama gobollada aan fiicneyn waxay isticmaalaan kaliya macluumaadka soo noqnoqoshada hoose - dhab ahaan dabeecadda lidka ku ah mipmaps ee sawirada caadiga ah.
Saamaynta Istiraatijiyadeed
Xawaaraha iyo miisaanka
Visual AI wuxuu si otomaatig ah u samayn karaa baadhista, ogaanshaha, iyo sumadaynta hawlaha miisaanka.
Xulashada dhismayaasha
Kooxaha hal-abuurka leh waxay hindise karaan fikradaha si dhakhso leh iyagoo leh dib-u-eegis buugeed yar.
Kooxda iyo socodka shaqada
Hawlgalladu waxay isticmaali karaan calaamadaha muuqaalka iyo muuqaalka kuwaas oo markii hore adkeyd in la farsameeyo.
Mustaqbalka Mip-NeRF iyo Goobaha Shucaaca ee Ka-hortagga ah
Mip-NeRF waxay soo saartay qoys ka kooban beero-ka-hortagga. Mip-NeRF 360 waxay fidisay koolooyinka bannaanka ee aan xuduudda lahayn ee leh xanuunka foosha, iyo Zip-NeRF waxay isku dartay ka-hortagga koontada-ku-saleysan ee leh matalo xashiish-degdegga ah si loo helo tayada iyo xawaaraha labadaba. Filo fikradda isku-dhafka ah ee niyad-jabka ah si ay ugu sii guurto Gaussian splating iyo dhuumaha-waqtiga-dhabta ah, halkaas oo qaab-badan, ku-bandhig-la'aanta taleefannada iyo dhegaha-madaxaha ay tahay yoolka AR, khariidaynta, iyo qabsashada immersive.
Dhaqangelinta Adduunka-dhabta ah
U soo bandhigida shay la qabtay si nadiif ah fiiryaha alaabta taas oo u ogolaanaysa isticmaalayaasha inay ka soo dhaweeyaan muuqaalka qolka buuxa ilaa tafaasiisha dusha sare iyada oo aan liiqliqin.
Dib-u-dhiska muuqaallo waaweyn oo bannaanka ah (iyada oo loo sii marayo Mip-NeRF 360) ee dalxiiska casriga ah iyo socod-ma-guurtada guryaha halkaasoo kamaradu ku dhex socoto qoto dheer oo ballaadhan.
Soo saarista sawiro tababar joogto ah oo xalal badan oo loogu talagalay robotics-ka ama jilayaasha darawalnimada iskeed u madaxbannaan.
Soo saarista muuqaalo muuqaal sheeko-sameeye ah oo qallafsan oo filim iyo horudhac VFX ah halkaas oo magac-u-dhigiddu ay jebinayso tallaalka.
Khatarta & Dariiqyada Ilaalada
Xuquuqda sawirka iyo ogolaanshaha waxay noqon kartaa khataro sharci ah haddii caddayntu aanay caddayn.
Waxqabadka moodeelku wuu ku kala duwanaan karaa iftiinka, tirakoobka, iyo deegaanka.
Wanaagga beenta ah waxa laga yaabaa inaan la dareemin ilaa xadka kalsoonida aan la kormeerin.
Qorshe Hawleedka Dhaqangelinta
Qeex shuruudaha aqbalida ee saxnaanta, dib u celinta, iyo kharashyada khaladka.
Ku tijaabi xogta ku habboon xaaladaha wax soo saarka dhabta ah.
Ku dar dib u eegis bini'aadamka si aad u hesho kalsoonida hoose ama saameeynta sare.
Lasoco moodeel dhaqaaqa oo dib u cusboonaysii kamarada ama xogta kaydinta ka dib.
Sii wad Sahaminta
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Mip-NeRF and Anti-Aliased Radiance Fields quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Hagaha xiga
Goobaha Shucaaca Neural
Su'aalaha soo noqnoqda
What is Mip-NeRF and Anti-Aliased Radiance Fields?
Mip-NeRF waxay hagaajisaa muuqalka blushka ah, ee jeexjeexan ee aafeeya NeRF-kii asalka ahaa marka aad soo bandhigto muuqaallo masaafo kala duwan ama xalal kala duwan ah. Tani waxay ku samaynaysaa iyada oo raadraacaysa koofiyada halkii ay ka heli lahaayeen fallaadho khafiif ah oo aan dhammaad lahayn, taasoo ka dhigaysa muuqaalka 3D mid ka dhigaya mid fiiqan oo degdeg ah si loo tababaro.
Qaab noocee ah ayuu Mip-NeRF ka dhex raadiyaa goobta halkii uu ka isticmaali lahaa shucaaca khafiifka ah ee uu isticmaalo NeRF-kii asalka ahaa?
Mip-NeRF waxa ay kaamirada ka soo tuurtaa koontada iyada oo loo marayo pixel kasta oo u kala qaybisa niyad jab, iyada oo soo qabata mugga pixel-ku dhab ahaantii daboolayo.
Waa maxay magaca Mip-NeRF's encoding ee bedelaya codaynta booska caadiga ah?
Mip-NeRF waxa ay soo bandhigtaa codayn meelaysan oo isku dhafan, kaas oo dejinaya gobol (qas-dhowrka Gaussian-qiyaasta) halkii uu ka ahaan lahaa hal dhibic.
Waa maxay sababta NeRF-ga asalka ahi u soo saaro aliasing marka masaafada kamaradu is beddesho?
Muunadda raajada khafiifka ah waxay iska indha tirtaa raad-reebka pixel-ka, markaa pixels dhow iyo ka fogba waxaa loola dhaqmaa si isku mid ah, taasoo keenta jaggies iyo liidashada.
Sidee Mip-NeRF u qiyaastaa gobolka ku jira jahwareerka koonka ah si loo codeeyo?
Jahawareer kasta waxaa lagu qiyaasaa Gaussian kala duwanaansho leh, codeyntuna waxay xisaabineysaa qiimaha la filayo ee sinusoids-ka Gaussian-kaas.
Maxaa ku dhacaya astaamaha codaynta-soo noqnoqda sare ee jahwareerka weyn?
Sababtoo ah sinusoids-ka soo noqnoqda aad ayey u kala duwan yihiin niyad jab weyn, isku darka Gaussian-ka waxay u kaxaysaa eber, taasoo siinaysa mipmap-sida lidka-aliasing.