SlideShare a Scribd company logo
NeRF: Representing Scenes as Neural
Radiance Fields for View Synthesis
연세대학교 영상 및 비디오 패턴 인식 연구실
이형민
2021.2.9
PR-302
Novel View Synthesis
Novel View Synthesis
uBullet-Time Effect
N대의 카메라를 통한 Input è 임의의 시점(視點)에서 본 대상의 모습을 합성
https://www.matthewtancik.com/nerf
Novel View Synthesis
uParallax Effect
3D Space에 대한 Analysis가 필수적!!
Pinhole Camera Model
(PR-228: GeoNet)
Pinhole Camera Model
uPinhole Camera Model
Pinhole Camera Model
uReal Length??
Pinhole Camera Model
uNormalized Plane
Depth=1인 평면 위로 모든 점을 Projection!!
Pinhole Camera Model
uIntrinsic Parameter
Camera 정보를 이용하여 pixel domain으로 projection
Pinhole Camera Model
uIntrinsic Parameter
Pinhole Camera Model
uExtrinsic Parameter
NeRF: Representing Scenes as Neural Radiance Fields
for View Synthesis
[ECCV 2020]
NeRF
uScenario of Novel View Synthesis
N개의 시점에서 찍은 2D Image è 임의의 시점에서 찍은 2D Image
NeRF
uNeural Radiance Field
3D data를 통째로 모델링하겠다!!
• 이미지: 각 위치에 pixel 값을 저장한 Table의 형태
• Neural Radiance Field: 각 위치 값을 input으로 주면 output으로 RGB 값을 연산하는 함수의 형태
NeRF
uNeural Radiance Field
• Input: 위치( −𝟏, 𝟏 𝟑
)와 보는 방향 (Unit Vector; 보는 방향에 따라서도 RGB 값이 달라질 수 있다.)
• Output: RGB 값과 Density (투명도의 역수 개념)
그 함수를 Deep Neural Net으로 정의하고, 주어진 이미지들로 학습하자!
NeRF
uRay
NeRF
uProjection
Pixel 값 = 한 Ray 위에 존재하는 Point들의 RGB 값들의 Weighted Sum!!
NeRF
uProjection
• Density가 클수록 Weight가 커야 한다!
• 그 지점을 가로막고 있는 점들의 Density의 합이 작을수록 Weight가 커야 한다!
NeRF
uNormalized Device Coordinate (NDC)
• Real World Scene을 Cube 형태의 Space로 변환
NeRF
uNormalized Device Coordinate (NDC)
NeRF
NeRF
uNormalized Device Coordinate (NDC)
NeRF
uNormalized Device Coordinate (NDC)
NeRF
uNormalized Device Coordinate (NDC)
NeRF
uNormalized Device Coordinate (NDC)
NeRF
uImplementation
Max, Nelson. "Optical models for direct volume rendering." IEEE Transactions on Visualization and Computer Graphics 1.2 (1995): 99-108.
NeRF
uImplementation
• Hierarchical Volume Sampling
• 앞의 방법을 통해 𝑵𝒄개의 𝒕를 뽑고, 이를 이용하여 Coarse Network를 학습시킨다.
• Coarse Network를 통해 얻은 Distribution(𝑻(𝒓))에 따라 𝑵𝒇개의 𝒕를 추가적으로 뽑는다.
• 총 𝑵𝒄 + 𝑵𝒇개의 Sample을 이용하여 Fine Network를 학습시킨다.
NeRF
uImplementation
• Positional Encoding
NeRF
uResults
https://www.matthewtancik.com/nerf
Q&A

More Related Content

PDF
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
taeseon ryu
 
PPTX
3D Gaussian Splatting
taeseon ryu
 
PDF
Neural Radiance Field
Dong Heon Cho
 
PDF
Image-to-Image Translation
Junho Kim
 
PDF
Photo-realistic Single Image Super-resolution using a Generative Adversarial ...
Hansol Kang
 
PDF
Super resolution-review
Woojin Jeong
 
PDF
Deformable DETR Review [CDM]
Dongmin Choi
 
PPTX
알기쉬운 Variational autoencoder
홍배 김
 
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
taeseon ryu
 
3D Gaussian Splatting
taeseon ryu
 
Neural Radiance Field
Dong Heon Cho
 
Image-to-Image Translation
Junho Kim
 
Photo-realistic Single Image Super-resolution using a Generative Adversarial ...
Hansol Kang
 
Super resolution-review
Woojin Jeong
 
Deformable DETR Review [CDM]
Dongmin Choi
 
알기쉬운 Variational autoencoder
홍배 김
 

What's hot (20)

PPTX
【DL輪読会】"Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"
Deep Learning JP
 
PDF
[DL輪読会]Convolutional Conditional Neural Processesと Neural Processes Familyの紹介
Deep Learning JP
 
PDF
Cvpr 2021 manydepth
Kenta Tanaka
 
PPTX
【DL輪読会】GET3D: A Generative Model of High Quality 3D Textured Shapes Learned f...
Deep Learning JP
 
PDF
【メタサーベイ】Neural Fields
cvpaper. challenge
 
PDF
論文紹介「PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet」
Naoya Chiba
 
PDF
【DL輪読会】Novel View Synthesis with Diffusion Models
Deep Learning JP
 
PDF
深層生成モデルを用いたマルチモーダル学習
Masahiro Suzuki
 
PDF
[DL輪読会]NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
Deep Learning JP
 
PDF
【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision
Deep Learning JP
 
PPTX
Swin Transformer (ICCV'21 Best Paper) を完璧に理解する資料
Yusuke Uchida
 
PDF
Optimizer入門&最新動向
Motokawa Tetsuya
 
PPTX
[DL輪読会]MetaFormer is Actually What You Need for Vision
Deep Learning JP
 
PPTX
近年のHierarchical Vision Transformer
Yusuke Uchida
 
PDF
ECCV読み会 "Materials for Masses: SVBRDF Acquisition with a Single Mobile Phone ...
Hajime Mihara
 
PDF
Generating Diverse High-Fidelity Images with VQ-VAE-2
harmonylab
 
PPTX
SfM Learner系単眼深度推定手法について
Ryutaro Yamauchi
 
PPTX
[DL輪読会]Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Deep Learning JP
 
PPTX
SLAM勉強会(PTAM)
Masaya Kaneko
 
PPTX
【DL輪読会】ViT + Self Supervised Learningまとめ
Deep Learning JP
 
【DL輪読会】"Instant Neural Graphics Primitives with a Multiresolution Hash Encoding"
Deep Learning JP
 
[DL輪読会]Convolutional Conditional Neural Processesと Neural Processes Familyの紹介
Deep Learning JP
 
Cvpr 2021 manydepth
Kenta Tanaka
 
【DL輪読会】GET3D: A Generative Model of High Quality 3D Textured Shapes Learned f...
Deep Learning JP
 
【メタサーベイ】Neural Fields
cvpaper. challenge
 
論文紹介「PointNetLK: Robust & Efficient Point Cloud Registration Using PointNet」
Naoya Chiba
 
【DL輪読会】Novel View Synthesis with Diffusion Models
Deep Learning JP
 
深層生成モデルを用いたマルチモーダル学習
Masahiro Suzuki
 
[DL輪読会]NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
Deep Learning JP
 
【DL輪読会】DINOv2: Learning Robust Visual Features without Supervision
Deep Learning JP
 
Swin Transformer (ICCV'21 Best Paper) を完璧に理解する資料
Yusuke Uchida
 
Optimizer入門&最新動向
Motokawa Tetsuya
 
[DL輪読会]MetaFormer is Actually What You Need for Vision
Deep Learning JP
 
近年のHierarchical Vision Transformer
Yusuke Uchida
 
ECCV読み会 "Materials for Masses: SVBRDF Acquisition with a Single Mobile Phone ...
Hajime Mihara
 
Generating Diverse High-Fidelity Images with VQ-VAE-2
harmonylab
 
SfM Learner系単眼深度推定手法について
Ryutaro Yamauchi
 
[DL輪読会]Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Deep Learning JP
 
SLAM勉強会(PTAM)
Masaya Kaneko
 
【DL輪読会】ViT + Self Supervised Learningまとめ
Deep Learning JP
 
Ad

Similar to PR-302: NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis (10)

PPTX
20230213_ComputerVision_연구.pptx
ssuser7807522
 
PDF
Neural Radiance Fields & Neural Rendering.pdf
NavneetPaul2
 
PPTX
Scene Representation Networks(NIPS 2019)_OJung
nemostar51
 
PDF
Single shot multiboxdetectors
지현 백
 
PDF
(Paper Review) Reconstruction of Monte Carlo Image Sequences using a Recurren...
MYEONGGYU LEE
 
PPTX
Single shot multiboxdetectors
지현 백
 
PPTX
Understanding neural radiance fields
Varun Bhaseen
 
PDF
TransNeRF
NavneetPaul2
 
PDF
[NDC2017] 딥러닝으로 게임 콘텐츠 제작하기 - VAE를 이용한 콘텐츠 생성 기법 연구 사례
Hwanhee Kim
 
PDF
PR-240: Modulating Image Restoration with Continual Levels via Adaptive Featu...
Hyeongmin Lee
 
20230213_ComputerVision_연구.pptx
ssuser7807522
 
Neural Radiance Fields & Neural Rendering.pdf
NavneetPaul2
 
Scene Representation Networks(NIPS 2019)_OJung
nemostar51
 
Single shot multiboxdetectors
지현 백
 
(Paper Review) Reconstruction of Monte Carlo Image Sequences using a Recurren...
MYEONGGYU LEE
 
Single shot multiboxdetectors
지현 백
 
Understanding neural radiance fields
Varun Bhaseen
 
TransNeRF
NavneetPaul2
 
[NDC2017] 딥러닝으로 게임 콘텐츠 제작하기 - VAE를 이용한 콘텐츠 생성 기법 연구 사례
Hwanhee Kim
 
PR-240: Modulating Image Restoration with Continual Levels via Adaptive Featu...
Hyeongmin Lee
 
Ad

More from Hyeongmin Lee (20)

PDF
PR-455: CoTracker: It is Better to Track Together
Hyeongmin Lee
 
PDF
PR-430: CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retri...
Hyeongmin Lee
 
PDF
PR-420: Scalable Model Compression by Entropy Penalized Reparameterization
Hyeongmin Lee
 
PDF
PR-409: Denoising Diffusion Probabilistic Models
Hyeongmin Lee
 
PDF
PR-395: Variational Image Compression with a Scale Hyperprior
Hyeongmin Lee
 
PDF
PR-386: Light Field Networks: Neural Scene Representations with Single-Evalua...
Hyeongmin Lee
 
PDF
PR-376: Softmax Splatting for Video Frame Interpolation
Hyeongmin Lee
 
PDF
PR-365: Fast object detection in compressed video
Hyeongmin Lee
 
PDF
PR-340: DVC: An End-to-end Deep Video Compression Framework
Hyeongmin Lee
 
PDF
PR-328: End-to-End Optimized Image Compression
Hyeongmin Lee
 
PDF
PR-315: Taming Transformers for High-Resolution Image Synthesis
Hyeongmin Lee
 
PDF
PR-278: RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
Hyeongmin Lee
 
PDF
Pr266
Hyeongmin Lee
 
PDF
PR-252: Making Convolutional Networks Shift-Invariant Again
Hyeongmin Lee
 
PDF
PR-228: Geonet: Unsupervised learning of dense depth, optical flow and camera...
Hyeongmin Lee
 
PDF
PR-214: FlowNet: Learning Optical Flow with Convolutional Networks
Hyeongmin Lee
 
PDF
[PR12] Making Convolutional Networks Shift-Invariant Again
Hyeongmin Lee
 
PPTX
Latest Frame interpolation Algorithms
Hyeongmin Lee
 
PPTX
[Paper Review] Temporal Generative Adversarial Nets with Singular Value Clipping
Hyeongmin Lee
 
PPTX
[Paper Review] A Middlebury Benchmark & Context-Aware Synthesis for Video Fra...
Hyeongmin Lee
 
PR-455: CoTracker: It is Better to Track Together
Hyeongmin Lee
 
PR-430: CLIP4Clip: An Empirical Study of CLIP for End to End Video Clip Retri...
Hyeongmin Lee
 
PR-420: Scalable Model Compression by Entropy Penalized Reparameterization
Hyeongmin Lee
 
PR-409: Denoising Diffusion Probabilistic Models
Hyeongmin Lee
 
PR-395: Variational Image Compression with a Scale Hyperprior
Hyeongmin Lee
 
PR-386: Light Field Networks: Neural Scene Representations with Single-Evalua...
Hyeongmin Lee
 
PR-376: Softmax Splatting for Video Frame Interpolation
Hyeongmin Lee
 
PR-365: Fast object detection in compressed video
Hyeongmin Lee
 
PR-340: DVC: An End-to-end Deep Video Compression Framework
Hyeongmin Lee
 
PR-328: End-to-End Optimized Image Compression
Hyeongmin Lee
 
PR-315: Taming Transformers for High-Resolution Image Synthesis
Hyeongmin Lee
 
PR-278: RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
Hyeongmin Lee
 
PR-252: Making Convolutional Networks Shift-Invariant Again
Hyeongmin Lee
 
PR-228: Geonet: Unsupervised learning of dense depth, optical flow and camera...
Hyeongmin Lee
 
PR-214: FlowNet: Learning Optical Flow with Convolutional Networks
Hyeongmin Lee
 
[PR12] Making Convolutional Networks Shift-Invariant Again
Hyeongmin Lee
 
Latest Frame interpolation Algorithms
Hyeongmin Lee
 
[Paper Review] Temporal Generative Adversarial Nets with Singular Value Clipping
Hyeongmin Lee
 
[Paper Review] A Middlebury Benchmark & Context-Aware Synthesis for Video Fra...
Hyeongmin Lee
 

Recently uploaded (20)

PDF
67243-Cooling and Heating & Calculation.pdf
DHAKA POLYTECHNIC
 
PDF
LEAP-1B presedntation xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
hatem173148
 
PPTX
FUNDAMENTALS OF ELECTRIC VEHICLES UNIT-1
MikkiliSuresh
 
PPTX
Module2 Data Base Design- ER and NF.pptx
gomathisankariv2
 
PDF
settlement FOR FOUNDATION ENGINEERS.pdf
Endalkazene
 
PDF
CAD-CAM U-1 Combined Notes_57761226_2025_04_22_14_40.pdf
shailendrapratap2002
 
PPTX
Inventory management chapter in automation and robotics.
atisht0104
 
PDF
Biodegradable Plastics: Innovations and Market Potential (www.kiu.ac.ug)
publication11
 
PDF
EVS+PRESENTATIONS EVS+PRESENTATIONS like
saiyedaqib429
 
PPTX
business incubation centre aaaaaaaaaaaaaa
hodeeesite4
 
PPT
Understanding the Key Components and Parts of a Drone System.ppt
Siva Reddy
 
PDF
top-5-use-cases-for-splunk-security-analytics.pdf
yaghutialireza
 
PDF
67243-Cooling and Heating & Calculation.pdf
DHAKA POLYTECHNIC
 
PPTX
22PCOAM21 Session 2 Understanding Data Source.pptx
Guru Nanak Technical Institutions
 
PDF
Advanced LangChain & RAG: Building a Financial AI Assistant with Real-Time Data
Soufiane Sejjari
 
PDF
Cryptography and Information :Security Fundamentals
Dr. Madhuri Jawale
 
PPTX
MULTI LEVEL DATA TRACKING USING COOJA.pptx
dollysharma12ab
 
PDF
20ME702-Mechatronics-UNIT-1,UNIT-2,UNIT-3,UNIT-4,UNIT-5, 2025-2026
Mohanumar S
 
PDF
STUDY OF NOVEL CHANNEL MATERIALS USING III-V COMPOUNDS WITH VARIOUS GATE DIEL...
ijoejnl
 
PDF
AI-Driven IoT-Enabled UAV Inspection Framework for Predictive Maintenance and...
ijcncjournal019
 
67243-Cooling and Heating & Calculation.pdf
DHAKA POLYTECHNIC
 
LEAP-1B presedntation xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
hatem173148
 
FUNDAMENTALS OF ELECTRIC VEHICLES UNIT-1
MikkiliSuresh
 
Module2 Data Base Design- ER and NF.pptx
gomathisankariv2
 
settlement FOR FOUNDATION ENGINEERS.pdf
Endalkazene
 
CAD-CAM U-1 Combined Notes_57761226_2025_04_22_14_40.pdf
shailendrapratap2002
 
Inventory management chapter in automation and robotics.
atisht0104
 
Biodegradable Plastics: Innovations and Market Potential (www.kiu.ac.ug)
publication11
 
EVS+PRESENTATIONS EVS+PRESENTATIONS like
saiyedaqib429
 
business incubation centre aaaaaaaaaaaaaa
hodeeesite4
 
Understanding the Key Components and Parts of a Drone System.ppt
Siva Reddy
 
top-5-use-cases-for-splunk-security-analytics.pdf
yaghutialireza
 
67243-Cooling and Heating & Calculation.pdf
DHAKA POLYTECHNIC
 
22PCOAM21 Session 2 Understanding Data Source.pptx
Guru Nanak Technical Institutions
 
Advanced LangChain & RAG: Building a Financial AI Assistant with Real-Time Data
Soufiane Sejjari
 
Cryptography and Information :Security Fundamentals
Dr. Madhuri Jawale
 
MULTI LEVEL DATA TRACKING USING COOJA.pptx
dollysharma12ab
 
20ME702-Mechatronics-UNIT-1,UNIT-2,UNIT-3,UNIT-4,UNIT-5, 2025-2026
Mohanumar S
 
STUDY OF NOVEL CHANNEL MATERIALS USING III-V COMPOUNDS WITH VARIOUS GATE DIEL...
ijoejnl
 
AI-Driven IoT-Enabled UAV Inspection Framework for Predictive Maintenance and...
ijcncjournal019
 

PR-302: NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis