Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
GNeRF: GAN-based Neural Radiance Field without ...
Search
peisuke
December 12, 2021
Technology
1
810
GNeRF: GAN-based Neural Radiance Field without Posed Camera
第9回全日本コンピュータビジョン勉強会
ICCV2021論文読み回発表資料
GNeRF: GAN-based Neural Radiance Field without Posed Camera
peisuke
December 12, 2021
Tweet
Share
More Decks by peisuke
See All by peisuke
VGGT: Visual Geometry Grounded Transformer
peisuke
1
1.2k
AI for Kids:小学生に画像認識を教えてみた話
peisuke
1
52
LangGraphで始めるマルチエージェントシステム
peisuke
14
4.7k
Self-RAG: Learning to Retrieve, Generate and Critique through Self-Reflections
peisuke
9
1.6k
Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields
peisuke
0
13k
LangChain Toolsの運用と改善
peisuke
5
2.9k
TTS Skins: Speaker Conversion via ASR
peisuke
0
420
A Quantum Computational Approach to Correspondence Problems on Point Sets
peisuke
0
750
F0-Consistent Many-to-many Non-parallel Voice Conversion via Conditional Autoencoder
peisuke
0
230
Other Decks in Technology
See All in Technology
GraphRAG グラフDBを使ったLLM生成(自作漫画DBを用いた具体例を用いて)
seaturt1e
1
170
AI駆動で進める依存ライブラリ更新 ─ Vue プロジェクトの品質向上と開発スピード改善の実践録
sayn0
1
360
ViteとTypeScriptのProject Referencesで 大規模モノレポのUIカタログのリリースサイクルを高速化する
shuta13
3
230
オブザーバビリティと育てた ID管理・認証認可基盤の歩み / The Journey of an ID Management, Authentication, and Authorization Platform Nurtured with Observability
kaminashi
2
1.5k
初海外がre:Inventだった人間の感じたこと
tommy0124
1
120
re:Inventに行くまでにやっておきたいこと
nagisa53
0
810
アノテーション作業書作成のGood Practice
cierpa0905
PRO
1
330
SRE × マネジメントレイヤーが挑戦した組織・会社のオブザーバビリティ改革 ― ビジネス価値と信頼性を両立するリアルな挑戦
coconala_engineer
0
320
AIの個性を理解し、指揮する
shoota
3
550
20251027_マルチエージェントとは
almondo_event
1
490
AI連携の新常識! 話題のMCPをはじめて学ぶ!
makoakiba
0
160
可観測性は開発環境から、開発環境にもオブザーバビリティ導入のススメ
layerx
PRO
4
2.3k
Featured
See All Featured
The Illustrated Children's Guide to Kubernetes
chrisshort
51
51k
How to Create Impact in a Changing Tech Landscape [PerfNow 2023]
tammyeverts
55
3k
Connecting the Dots Between Site Speed, User Experience & Your Business [WebExpo 2025]
tammyeverts
10
630
A Tale of Four Properties
chriscoyier
161
23k
Easily Structure & Communicate Ideas using Wireframe
afnizarnur
194
16k
Understanding Cognitive Biases in Performance Measurement
bluesmoon
31
2.7k
The Straight Up "How To Draw Better" Workshop
denniskardys
239
140k
Building a Scalable Design System with Sketch
lauravandoore
463
33k
A better future with KSS
kneath
239
18k
10 Git Anti Patterns You Should be Aware of
lemiorhan
PRO
658
61k
Chrome DevTools: State of the Union 2024 - Debugging React & Beyond
addyosmani
9
940
Let's Do A Bunch of Simple Stuff to Make Websites Faster
chriscoyier
508
140k
Transcript
GNeRF: GAN-based Neural Radiance Field without Posed Camera @peisuke 9
ABEJA Twitter @peisuke Github https://github.com/peisuke Qiita https://qiita.com/peisuke SlideShare https://www.slideshare.net/FujimotoKeisuke
• • GNeRF: GAN-based Neural Radiance Field without Posed Camera
• • NeRF • • • GAN Radiance Field • RF
NeRF • / 5 •
• • (x,y,z)T (φ1,θ1)T (φ2,θ2)T (x,y,z) φ2,θ2 (x,y,z) φ1,θ1
NeRF • • •
NeRF • NeRF in the Wild • NeRF • iNeRF
• iNeRF NeRF in the Wild BP
条件 iNeRF NeRF-- IDR GNeRF(提案⼿法) 事前学習 必要 不要 不要 不要
粗いカメラ座標 必要 必要 必要 不要 前⽅のみの制約 不要 必要 不要 不要 マスク 不要 不要 必要 不要 • •
• • • • • GAN Radiance Field • RF
• Phase A Phase B
RF • D E 3
GAN Real • • Discriminator Real • Dynamic Patch Sampling
/ 16x16 • Dynamic Patch Sampling
GAN Fake • • NeRF 16x16 • G
• 6 𝑹 = 𝑟!! 𝑟!" 𝑟!# 𝑟"! 𝑟"" 𝑟"#
𝑟#! 𝑟#" 𝑟## 𝒓 = 𝑟!! 𝑟!" 𝑟!# 𝑟"! 𝑟"" 𝑟"# $ • SO(3) NN • On the Continuity of Rotation Representations in Neural Networks
• Inversion network • Static Patch Sampling • E •
Static Patch Sampling
G G D G
• • Inversion network • refinement
• Refinement
• • • • • Photometric loss
• • Generator • NeRF • Discriminator • Instance Normalization
Spectral Normalization • Inversion network • Vision Transformer • • A • A B 1 iteration • B
• • COLMAP NeRF C+n • Refine C+r • Ground
Truth NeRF G+n • • Synthetic-NeRF image size: 400x400 DTU 500x400 • 100 • • Peak Signal-to-Noise Ratio SSIM LPIPS • / • Single TAITAN GPU 30
COLMAP •
• COLMAP Drums
難しいシーン 特徴の少ない シーン
IDR • 3D IDR • IDR • PNSR •
• COLMAP • Drums Lego COLMAP
Ablation Study • • GAN InvNet. PhaseB • • A
B A ABABABAB B
• • • COLMAP • •
• NeRF • • GAN RF • RF • •
• COLMAP