SlideShare ist ein Scribd-Unternehmen logo
1 von 37
Semantic Segmentation on
Satellite Imagery
Rahul Bhojwani, Nina Domingo,
Benjamin Mayhew, Christy Tsz-En Wang
Kaggle: Can you train an eye in the sky?
Challenge: The Defence Science and
Technology Laboratory (DSTL) is seeking
novel solutions to alleviate the burden on
their image analysts and challenges
kagglers to accurately identify and classify
objects in overhead satellite imagery.
Introduction Data Methods Results
What’s in a picture?
Introduction Data Methods Results
How is this useful?
Medical imaging Agriculture Surveillance
Introduction Data Methods Results
Data
Input: 25 1km x 1km satellite images in both 3-band and 16-band
formats
● Format: GeoTiff
● Images are taken from the same region but coordinates are
transformed so the location is obscured
Object class: every class is provided in the form of a Multipolygon
● Format: Geojson or WKT
Introduction Data Methods Results
Object Class Types
Buildings Crops
Misc. Manmade Structures Waterway
Roads Standing Water
Track Vehicle Large
Trees Vehicle Small
Introduction Data Methods Results
Data Processing of Labels
Introduction Data Methods Results
Match [0,1] coordinates to
pixel coordinates
Compute projection factors for
multipolygon
Data Processing of Labels
Introduction Data Methods Results
Multipolygons to shapely objects
Project geometry to pixel coordinates
Shapely objects to shapefiles
to tiff files
Data Processing
Original image Object mask Superimposed image
Introduction Data Methods Results
Introduction Data Methods Results
Object Class Type Distribution
Introduction Data Methods Results
Average Number of Polygons Distribution
Introduction Data Methods Results
More Data Processing
25 512x512
images
Introduction Data Methods Results
25 ~3300x3300
images
25 3072x3072
images
900 512x512
images
DIRECT SCALING PARTITION
Methods - Semantic Segmentation with Deep Learning
Important deep learning
models for semantic
segmentation:
● Fully Convolutional
Network [Nov 2014]
● U-net [May 2015]
● Segnet [Nov 2015]
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
VGG-16:
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
Introduction Data Methods Results
Fully Convolutional
Network:
● No fully
connected
● Skip
connection
● VGG-16
Methods - Semantic Segmentation with Deep Learning
U-Net:
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
U-Net:
● Encoder-Decoder network.
● Every decoding phase is convolved with trainable filters.
● Copy the encoder embedding to the corresponding decoder.
● Data Augmentation [Stretching and rotation].
● Weighted Cross Entropy.
● Forces network to learn the border pixels.
Introduction Data Methods Results
Methods - Encode/Contracting path
Goal:
● Retain context and
localization accuracy.
Operations:
● Convolution
● Non Linearity (ReLU)
● Pooling
● But skip the fully connected
layers
Introduction Data Methods Results
3x3 Convolution with
no padding, stride of 2
Methods - Semantic Segmentation with Deep Learning
Segnet Architecture:
Introduction Data Methods Results
Methods - Decode/Expansive path
Goal:
● To recover the object details and
spatial dimension
Operation:
● “Up-convolution”/ “upsampling”
● Concatenate with the corresponding
cropped encoder feature maps
● Convolution layers
● ReLU
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
Segnet:
● Encoding part is exactly VGG-16
● Use Trained weights from VGG-16 [Excluding the last fully connected
layer]
● Decoder uses the pooling indices from max pooling step of
corresponding encoder.
● The upsampled maps were convolved with trainable filters.
● Unlike U-Net they don’t copy the entire encoding.
● Reduced the trainable parameters from 134M → 14.7M
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
Segnet Unpooling:
Introduction Data Methods Results
Methods - Semantic Segmentation with Deep Learning
FCN vs Segnet:
Introduction Data Methods Results
Training U-net
Pixel-wise soft-max + cross entropy loss function
Methods: How does upsampling work?
Transposed convolution (fractionally strided
convolution/deconvolution)
● Reconstructs the spatial resolution
● The weights are learnable
● It is NOT reverse convolution process
Introduction Data Methods Results
Transposed 2x2 convolution
with no padding, stride of 2 and
kernel of 3
Convolution as matrix
multiplication
4 x 4
3 x 3
Convolution as matrix
multiplication
4 x 16
16 x 1
4 x 1
Transposed convolution as
matrix multiplication
(16 x 4) (4 x 1) = (16 x 1)
● Dimension of input and output swap
● Uses transpose of convolution matrix
Preliminary results: partitioned images [900x512x512]
Introduction Data Methods Results
Epoch Loss Acc Epoch Loss Acc
1 0.2356 0.9587 6 NA NA
2 0.1763 0.9587 7 NA NA
3 ETA: ~1 day 8 NA NA
4 NA NA 9 NA NA
5 NA NA 10 NA NA
Next Steps
Actual Next Steps:
▫ Include more classes as part of our training.
▫ Tuning the hyperparameters of the model.
▫ Making the segnet work.
Future Works:
▫ Exploring more recently published models. Eg: Deeplab
v3[2018]
▫ Use higher computing resources to run the models
faster.
References:
▫ Ronneberger, O. (2017). Invited Talk: U-Net Convolutional Networks for Biomedical Image Segmentation.
Informatik Aktuell Bildverarbeitung Für Die Medizin 2017, 3-3. doi:10.1007/978-3-662-54345-0_3
▫ Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. 2015 IEEE
Conference on Computer Vision and Pattern Recognition (CVPR). doi:10.1109/cvpr.2015.7298965
▫ Badrinarayanan, V., Kendall, A., & Cipolla, R. (2017). SegNet: A Deep Convolutional Encoder-Decoder
Architecture for Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12),
2481-2495. doi:10.1109/tpami.2016.264461
▫ https://towardsdatascience.com/types-of-convolutions-in-deep-learning-717013397f4d
▫ https://www.cs.toronto.edu/~frossard/post/vgg16/
▫ https://medium.com/@wilburdes/semantic-segmentation-using-fully-convolutional-neural-networks-
86e45336f99b
▫ https://www.kaggle.com/c/dstl-satellite-imagery-feature-detection
Questions?
Extras:
Methods: dilated/atrous convolutions
Goal:
● Take away need to pool layers
Operations:
● Apply predefined gaps between each pixel
of input image
● Replace pooling layer from pretrained
classification system with dilated
convolution
e.g. 2-dilated convolution
Introduction Data Methods Results
Kaggle: Evaluation
Average Jaccard Index between the predicted multipolygons and actual
multipolygons. The Jaccard Index for two regions is the ratio of the area of the
intersection to the area of the union.
Jaccard =TP/(TP + FP + FN) = |A∩B|/|A∪B| = |A∩B|/(|A|+|B|−|A∩B|)
Introduction Data Methods Results

Weitere ähnliche Inhalte

Was ist angesagt?

Object detection
Object detectionObject detection
Object detectionSomesh Vyas
 
Object detection with deep learning
Object detection with deep learningObject detection with deep learning
Object detection with deep learningSushant Shrivastava
 
Landuse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningLanduse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningDataWorks Summit
 
Machine learning in image processing
Machine learning in image processingMachine learning in image processing
Machine learning in image processingData Science Thailand
 
Deep Learning - Convolutional Neural Networks
Deep Learning - Convolutional Neural NetworksDeep Learning - Convolutional Neural Networks
Deep Learning - Convolutional Neural NetworksChristian Perone
 
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation...
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation..."Semantic Segmentation for Scene Understanding: Algorithms and Implementation...
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation...Edge AI and Vision Alliance
 
Statistical Pattern recognition(1)
Statistical Pattern recognition(1)Statistical Pattern recognition(1)
Statistical Pattern recognition(1)Syed Atif Naseem
 
Spatial Filters (Digital Image Processing)
Spatial Filters (Digital Image Processing)Spatial Filters (Digital Image Processing)
Spatial Filters (Digital Image Processing)Kalyan Acharjya
 
Image segmentation with deep learning
Image segmentation with deep learningImage segmentation with deep learning
Image segmentation with deep learningAntonio Rueda-Toicen
 
Dimensionality Reduction
Dimensionality ReductionDimensionality Reduction
Dimensionality Reductionmrizwan969
 
Real Time Object Tracking
Real Time Object TrackingReal Time Object Tracking
Real Time Object TrackingVanya Valindria
 
Fundamental steps in image processing
Fundamental steps in image processingFundamental steps in image processing
Fundamental steps in image processingPremaPRC211300301103
 
Introduction to object detection
Introduction to object detectionIntroduction to object detection
Introduction to object detectionBrodmann17
 
Chapter 4 Image Processing: Image Transformation
Chapter 4 Image Processing: Image TransformationChapter 4 Image Processing: Image Transformation
Chapter 4 Image Processing: Image TransformationVarun Ojha
 

Was ist angesagt? (20)

Psuedo color
Psuedo colorPsuedo color
Psuedo color
 
Object detection
Object detectionObject detection
Object detection
 
Object detection with deep learning
Object detection with deep learningObject detection with deep learning
Object detection with deep learning
 
Landuse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep LearningLanduse Classification from Satellite Imagery using Deep Learning
Landuse Classification from Satellite Imagery using Deep Learning
 
Machine learning in image processing
Machine learning in image processingMachine learning in image processing
Machine learning in image processing
 
Deep learning
Deep learningDeep learning
Deep learning
 
Deep Learning - Convolutional Neural Networks
Deep Learning - Convolutional Neural NetworksDeep Learning - Convolutional Neural Networks
Deep Learning - Convolutional Neural Networks
 
Object detection
Object detectionObject detection
Object detection
 
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation...
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation..."Semantic Segmentation for Scene Understanding: Algorithms and Implementation...
"Semantic Segmentation for Scene Understanding: Algorithms and Implementation...
 
Computer vision
Computer vision Computer vision
Computer vision
 
Statistical Pattern recognition(1)
Statistical Pattern recognition(1)Statistical Pattern recognition(1)
Statistical Pattern recognition(1)
 
Spatial Filters (Digital Image Processing)
Spatial Filters (Digital Image Processing)Spatial Filters (Digital Image Processing)
Spatial Filters (Digital Image Processing)
 
Image segmentation with deep learning
Image segmentation with deep learningImage segmentation with deep learning
Image segmentation with deep learning
 
Dimensionality Reduction
Dimensionality ReductionDimensionality Reduction
Dimensionality Reduction
 
Real Time Object Tracking
Real Time Object TrackingReal Time Object Tracking
Real Time Object Tracking
 
Object tracking
Object trackingObject tracking
Object tracking
 
Fundamental steps in image processing
Fundamental steps in image processingFundamental steps in image processing
Fundamental steps in image processing
 
Introduction to object detection
Introduction to object detectionIntroduction to object detection
Introduction to object detection
 
Principal component analysis
Principal component analysisPrincipal component analysis
Principal component analysis
 
Chapter 4 Image Processing: Image Transformation
Chapter 4 Image Processing: Image TransformationChapter 4 Image Processing: Image Transformation
Chapter 4 Image Processing: Image Transformation
 

Ähnlich wie Semantic Segmentation on Satellite Imagery

Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)
Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)
Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)Universitat Politècnica de Catalunya
 
FastV2C-HandNet - ICICC 2020
FastV2C-HandNet - ICICC 2020FastV2C-HandNet - ICICC 2020
FastV2C-HandNet - ICICC 2020RohanLekhwani
 
Deep image retrieval learning global representations for image search
Deep image retrieval  learning global representations for image searchDeep image retrieval  learning global representations for image search
Deep image retrieval learning global representations for image searchUniversitat Politècnica de Catalunya
 
Deep image retrieval - learning global representations for image search - ub ...
Deep image retrieval - learning global representations for image search - ub ...Deep image retrieval - learning global representations for image search - ub ...
Deep image retrieval - learning global representations for image search - ub ...Universitat de Barcelona
 
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...MLconf
 
Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
Point-GNN: Graph Neural Network for 3D Object Detection in a Point CloudPoint-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
Point-GNN: Graph Neural Network for 3D Object Detection in a Point CloudNuwan Sriyantha Bandara
 
DDGK: Learning Graph Representations for Deep Divergence Graph Kernels
DDGK: Learning Graph Representations for Deep Divergence Graph KernelsDDGK: Learning Graph Representations for Deep Divergence Graph Kernels
DDGK: Learning Graph Representations for Deep Divergence Graph Kernelsivaderivader
 
Survey on optical flow estimation with DL
Survey on optical flow estimation with DLSurvey on optical flow estimation with DL
Survey on optical flow estimation with DLLeapMind Inc
 
Heterogeneous data fusion with multiple kernel growing self organizing maps
Heterogeneous data fusion with multiple kernel growing self organizing mapsHeterogeneous data fusion with multiple kernel growing self organizing maps
Heterogeneous data fusion with multiple kernel growing self organizing mapsPruthuvi Maheshakya Wijewardena
 
物件偵測與辨識技術
物件偵測與辨識技術物件偵測與辨識技術
物件偵測與辨識技術CHENHuiMei
 
Low complexity features for jpeg steganalysis using undecimated dct
Low complexity features for jpeg steganalysis using undecimated dctLow complexity features for jpeg steganalysis using undecimated dct
Low complexity features for jpeg steganalysis using undecimated dctPvrtechnologies Nellore
 
Pyramid Vector Quantization
Pyramid Vector QuantizationPyramid Vector Quantization
Pyramid Vector QuantizationShahDhruv21
 
Garbage Classification Using Deep Learning Techniques
Garbage Classification Using Deep Learning TechniquesGarbage Classification Using Deep Learning Techniques
Garbage Classification Using Deep Learning TechniquesIRJET Journal
 
How might machine learning help advance solar PV research?
How might machine learning help advance solar PV research?How might machine learning help advance solar PV research?
How might machine learning help advance solar PV research?Anubhav Jain
 
Traffic Sign Recognition System
Traffic Sign Recognition SystemTraffic Sign Recognition System
Traffic Sign Recognition SystemIRJET Journal
 
DALL-E.pdf
DALL-E.pdfDALL-E.pdf
DALL-E.pdfdsfajkh
 
IRJET- 3D Object Recognition of Car Image Detection
IRJET-  	  3D Object Recognition of Car Image DetectionIRJET-  	  3D Object Recognition of Car Image Detection
IRJET- 3D Object Recognition of Car Image DetectionIRJET Journal
 
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...Seiya Ito
 

Ähnlich wie Semantic Segmentation on Satellite Imagery (20)

Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)
Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)
Image Segmentation (D3L1 2017 UPC Deep Learning for Computer Vision)
 
FastV2C-HandNet - ICICC 2020
FastV2C-HandNet - ICICC 2020FastV2C-HandNet - ICICC 2020
FastV2C-HandNet - ICICC 2020
 
Deep image retrieval learning global representations for image search
Deep image retrieval  learning global representations for image searchDeep image retrieval  learning global representations for image search
Deep image retrieval learning global representations for image search
 
Deep image retrieval - learning global representations for image search - ub ...
Deep image retrieval - learning global representations for image search - ub ...Deep image retrieval - learning global representations for image search - ub ...
Deep image retrieval - learning global representations for image search - ub ...
 
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
Hanjun Dai, PhD Student, School of Computational Science and Engineering, Geo...
 
Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
Point-GNN: Graph Neural Network for 3D Object Detection in a Point CloudPoint-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
Point-GNN: Graph Neural Network for 3D Object Detection in a Point Cloud
 
DDGK: Learning Graph Representations for Deep Divergence Graph Kernels
DDGK: Learning Graph Representations for Deep Divergence Graph KernelsDDGK: Learning Graph Representations for Deep Divergence Graph Kernels
DDGK: Learning Graph Representations for Deep Divergence Graph Kernels
 
Survey on optical flow estimation with DL
Survey on optical flow estimation with DLSurvey on optical flow estimation with DL
Survey on optical flow estimation with DL
 
Heterogeneous data fusion with multiple kernel growing self organizing maps
Heterogeneous data fusion with multiple kernel growing self organizing mapsHeterogeneous data fusion with multiple kernel growing self organizing maps
Heterogeneous data fusion with multiple kernel growing self organizing maps
 
PointNet
PointNetPointNet
PointNet
 
物件偵測與辨識技術
物件偵測與辨識技術物件偵測與辨識技術
物件偵測與辨識技術
 
Low complexity features for jpeg steganalysis using undecimated dct
Low complexity features for jpeg steganalysis using undecimated dctLow complexity features for jpeg steganalysis using undecimated dct
Low complexity features for jpeg steganalysis using undecimated dct
 
Pyramid Vector Quantization
Pyramid Vector QuantizationPyramid Vector Quantization
Pyramid Vector Quantization
 
Garbage Classification Using Deep Learning Techniques
Garbage Classification Using Deep Learning TechniquesGarbage Classification Using Deep Learning Techniques
Garbage Classification Using Deep Learning Techniques
 
How might machine learning help advance solar PV research?
How might machine learning help advance solar PV research?How might machine learning help advance solar PV research?
How might machine learning help advance solar PV research?
 
Traffic Sign Recognition System
Traffic Sign Recognition SystemTraffic Sign Recognition System
Traffic Sign Recognition System
 
DALL-E.pdf
DALL-E.pdfDALL-E.pdf
DALL-E.pdf
 
IRJET- 3D Object Recognition of Car Image Detection
IRJET-  	  3D Object Recognition of Car Image DetectionIRJET-  	  3D Object Recognition of Car Image Detection
IRJET- 3D Object Recognition of Car Image Detection
 
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...
[3D勉強会@関東] Deep Reinforcement Learning of Volume-guided Progressive View Inpa...
 
Deep 3D Visual Analysis - Javier Ruiz-Hidalgo - UPC Barcelona 2017
Deep 3D Visual Analysis - Javier Ruiz-Hidalgo - UPC Barcelona 2017Deep 3D Visual Analysis - Javier Ruiz-Hidalgo - UPC Barcelona 2017
Deep 3D Visual Analysis - Javier Ruiz-Hidalgo - UPC Barcelona 2017
 

Kürzlich hochgeladen

04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationshipsccctableauusergroup
 
Mature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxMature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxolyaivanovalion
 
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...Suhani Kapoor
 
BigBuy dropshipping via API with DroFx.pptx
BigBuy dropshipping via API with DroFx.pptxBigBuy dropshipping via API with DroFx.pptx
BigBuy dropshipping via API with DroFx.pptxolyaivanovalion
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxolyaivanovalion
 
Ravak dropshipping via API with DroFx.pptx
Ravak dropshipping via API with DroFx.pptxRavak dropshipping via API with DroFx.pptx
Ravak dropshipping via API with DroFx.pptxolyaivanovalion
 
Week-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interactionWeek-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interactionfulawalesam
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfRachmat Ramadhan H
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxfirstjob4
 
Schema on read is obsolete. Welcome metaprogramming..pdf
Schema on read is obsolete. Welcome metaprogramming..pdfSchema on read is obsolete. Welcome metaprogramming..pdf
Schema on read is obsolete. Welcome metaprogramming..pdfLars Albertsson
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysismanisha194592
 
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改atducpo
 
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Callshivangimorya083
 
RA-11058_IRR-COMPRESS Do 198 series of 1998
RA-11058_IRR-COMPRESS Do 198 series of 1998RA-11058_IRR-COMPRESS Do 198 series of 1998
RA-11058_IRR-COMPRESS Do 198 series of 1998YohFuh
 
B2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxB2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxStephen266013
 
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service Amravati
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service AmravatiVIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service Amravati
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service AmravatiSuhani Kapoor
 
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...Delhi Call girls
 
Invezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz1
 

Kürzlich hochgeladen (20)

꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
꧁❤ Aerocity Call Girls Service Aerocity Delhi ❤꧂ 9999965857 ☎️ Hard And Sexy ...
 
04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships04242024_CCC TUG_Joins and Relationships
04242024_CCC TUG_Joins and Relationships
 
Mature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptxMature dropshipping via API with DroFx.pptx
Mature dropshipping via API with DroFx.pptx
 
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...
VIP High Profile Call Girls Amravati Aarushi 8250192130 Independent Escort Se...
 
BigBuy dropshipping via API with DroFx.pptx
BigBuy dropshipping via API with DroFx.pptxBigBuy dropshipping via API with DroFx.pptx
BigBuy dropshipping via API with DroFx.pptx
 
Midocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFxMidocean dropshipping via API with DroFx
Midocean dropshipping via API with DroFx
 
Ravak dropshipping via API with DroFx.pptx
Ravak dropshipping via API with DroFx.pptxRavak dropshipping via API with DroFx.pptx
Ravak dropshipping via API with DroFx.pptx
 
Week-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interactionWeek-01-2.ppt BBB human Computer interaction
Week-01-2.ppt BBB human Computer interaction
 
Sampling (random) method and Non random.ppt
Sampling (random) method and Non random.pptSampling (random) method and Non random.ppt
Sampling (random) method and Non random.ppt
 
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdfMarket Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
Market Analysis in the 5 Largest Economic Countries in Southeast Asia.pdf
 
Introduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptxIntroduction-to-Machine-Learning (1).pptx
Introduction-to-Machine-Learning (1).pptx
 
Schema on read is obsolete. Welcome metaprogramming..pdf
Schema on read is obsolete. Welcome metaprogramming..pdfSchema on read is obsolete. Welcome metaprogramming..pdf
Schema on read is obsolete. Welcome metaprogramming..pdf
 
April 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's AnalysisApril 2024 - Crypto Market Report's Analysis
April 2024 - Crypto Market Report's Analysis
 
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改
代办国外大学文凭《原版美国UCLA文凭证书》加州大学洛杉矶分校毕业证制作成绩单修改
 
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
꧁❤ Greater Noida Call Girls Delhi ❤꧂ 9711199171 ☎️ Hard And Sexy Vip Call
 
RA-11058_IRR-COMPRESS Do 198 series of 1998
RA-11058_IRR-COMPRESS Do 198 series of 1998RA-11058_IRR-COMPRESS Do 198 series of 1998
RA-11058_IRR-COMPRESS Do 198 series of 1998
 
B2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docxB2 Creative Industry Response Evaluation.docx
B2 Creative Industry Response Evaluation.docx
 
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service Amravati
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service AmravatiVIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service Amravati
VIP Call Girls in Amravati Aarohi 8250192130 Independent Escort Service Amravati
 
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...
Call Girls in Sarai Kale Khan Delhi 💯 Call Us 🔝9205541914 🔝( Delhi) Escorts S...
 
Invezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signalsInvezz.com - Grow your wealth with trading signals
Invezz.com - Grow your wealth with trading signals
 

Semantic Segmentation on Satellite Imagery

  • 1. Semantic Segmentation on Satellite Imagery Rahul Bhojwani, Nina Domingo, Benjamin Mayhew, Christy Tsz-En Wang
  • 2. Kaggle: Can you train an eye in the sky? Challenge: The Defence Science and Technology Laboratory (DSTL) is seeking novel solutions to alleviate the burden on their image analysts and challenges kagglers to accurately identify and classify objects in overhead satellite imagery. Introduction Data Methods Results
  • 3. What’s in a picture? Introduction Data Methods Results
  • 4. How is this useful? Medical imaging Agriculture Surveillance Introduction Data Methods Results
  • 5. Data Input: 25 1km x 1km satellite images in both 3-band and 16-band formats ● Format: GeoTiff ● Images are taken from the same region but coordinates are transformed so the location is obscured Object class: every class is provided in the form of a Multipolygon ● Format: Geojson or WKT Introduction Data Methods Results
  • 6. Object Class Types Buildings Crops Misc. Manmade Structures Waterway Roads Standing Water Track Vehicle Large Trees Vehicle Small Introduction Data Methods Results
  • 7. Data Processing of Labels Introduction Data Methods Results Match [0,1] coordinates to pixel coordinates Compute projection factors for multipolygon
  • 8. Data Processing of Labels Introduction Data Methods Results Multipolygons to shapely objects Project geometry to pixel coordinates Shapely objects to shapefiles to tiff files
  • 9. Data Processing Original image Object mask Superimposed image Introduction Data Methods Results
  • 11. Object Class Type Distribution Introduction Data Methods Results
  • 12. Average Number of Polygons Distribution Introduction Data Methods Results
  • 13. More Data Processing 25 512x512 images Introduction Data Methods Results 25 ~3300x3300 images 25 3072x3072 images 900 512x512 images DIRECT SCALING PARTITION
  • 14. Methods - Semantic Segmentation with Deep Learning Important deep learning models for semantic segmentation: ● Fully Convolutional Network [Nov 2014] ● U-net [May 2015] ● Segnet [Nov 2015] Introduction Data Methods Results
  • 15. Methods - Semantic Segmentation with Deep Learning VGG-16: Introduction Data Methods Results
  • 16. Methods - Semantic Segmentation with Deep Learning Introduction Data Methods Results Fully Convolutional Network: ● No fully connected ● Skip connection ● VGG-16
  • 17. Methods - Semantic Segmentation with Deep Learning U-Net: Introduction Data Methods Results
  • 18. Methods - Semantic Segmentation with Deep Learning U-Net: ● Encoder-Decoder network. ● Every decoding phase is convolved with trainable filters. ● Copy the encoder embedding to the corresponding decoder. ● Data Augmentation [Stretching and rotation]. ● Weighted Cross Entropy. ● Forces network to learn the border pixels. Introduction Data Methods Results
  • 19. Methods - Encode/Contracting path Goal: ● Retain context and localization accuracy. Operations: ● Convolution ● Non Linearity (ReLU) ● Pooling ● But skip the fully connected layers Introduction Data Methods Results 3x3 Convolution with no padding, stride of 2
  • 20. Methods - Semantic Segmentation with Deep Learning Segnet Architecture: Introduction Data Methods Results
  • 21. Methods - Decode/Expansive path Goal: ● To recover the object details and spatial dimension Operation: ● “Up-convolution”/ “upsampling” ● Concatenate with the corresponding cropped encoder feature maps ● Convolution layers ● ReLU Introduction Data Methods Results
  • 22. Methods - Semantic Segmentation with Deep Learning Segnet: ● Encoding part is exactly VGG-16 ● Use Trained weights from VGG-16 [Excluding the last fully connected layer] ● Decoder uses the pooling indices from max pooling step of corresponding encoder. ● The upsampled maps were convolved with trainable filters. ● Unlike U-Net they don’t copy the entire encoding. ● Reduced the trainable parameters from 134M → 14.7M Introduction Data Methods Results
  • 23. Methods - Semantic Segmentation with Deep Learning Segnet Unpooling: Introduction Data Methods Results
  • 24. Methods - Semantic Segmentation with Deep Learning FCN vs Segnet: Introduction Data Methods Results
  • 25. Training U-net Pixel-wise soft-max + cross entropy loss function
  • 26. Methods: How does upsampling work? Transposed convolution (fractionally strided convolution/deconvolution) ● Reconstructs the spatial resolution ● The weights are learnable ● It is NOT reverse convolution process Introduction Data Methods Results Transposed 2x2 convolution with no padding, stride of 2 and kernel of 3
  • 29. Transposed convolution as matrix multiplication (16 x 4) (4 x 1) = (16 x 1) ● Dimension of input and output swap ● Uses transpose of convolution matrix
  • 30. Preliminary results: partitioned images [900x512x512] Introduction Data Methods Results Epoch Loss Acc Epoch Loss Acc 1 0.2356 0.9587 6 NA NA 2 0.1763 0.9587 7 NA NA 3 ETA: ~1 day 8 NA NA 4 NA NA 9 NA NA 5 NA NA 10 NA NA
  • 32. Actual Next Steps: ▫ Include more classes as part of our training. ▫ Tuning the hyperparameters of the model. ▫ Making the segnet work. Future Works: ▫ Exploring more recently published models. Eg: Deeplab v3[2018] ▫ Use higher computing resources to run the models faster.
  • 33. References: ▫ Ronneberger, O. (2017). Invited Talk: U-Net Convolutional Networks for Biomedical Image Segmentation. Informatik Aktuell Bildverarbeitung Für Die Medizin 2017, 3-3. doi:10.1007/978-3-662-54345-0_3 ▫ Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). doi:10.1109/cvpr.2015.7298965 ▫ Badrinarayanan, V., Kendall, A., & Cipolla, R. (2017). SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(12), 2481-2495. doi:10.1109/tpami.2016.264461 ▫ https://towardsdatascience.com/types-of-convolutions-in-deep-learning-717013397f4d ▫ https://www.cs.toronto.edu/~frossard/post/vgg16/ ▫ https://medium.com/@wilburdes/semantic-segmentation-using-fully-convolutional-neural-networks- 86e45336f99b ▫ https://www.kaggle.com/c/dstl-satellite-imagery-feature-detection
  • 36. Methods: dilated/atrous convolutions Goal: ● Take away need to pool layers Operations: ● Apply predefined gaps between each pixel of input image ● Replace pooling layer from pretrained classification system with dilated convolution e.g. 2-dilated convolution Introduction Data Methods Results
  • 37. Kaggle: Evaluation Average Jaccard Index between the predicted multipolygons and actual multipolygons. The Jaccard Index for two regions is the ratio of the area of the intersection to the area of the union. Jaccard =TP/(TP + FP + FN) = |A∩B|/|A∪B| = |A∩B|/(|A|+|B|−|A∩B|) Introduction Data Methods Results

Hinweis der Redaktion

  1. In December 2016, Kaggle hosted a 3-month competition in which the UK’s...
  2. But why try to do this? Medical imaging: detect location of a tumor Improve precision agriculture, identify plant disease General surveillance purposes
  3. For this specific challenge, we were provided with…. Multipolygon is a collection of polygons and these polygons represent objects in an image
  4. There are 10 types of object classes kagglers were challenged to identify...
  5. We also wanted to show you a video of the what the different object masks look like when superimposed to the original image...
  6. We also did a quick analysis of our object class distribution...
  7. I also mentioned that our object masks are provided in the form of multipolygons… A multipolygon of trees is made of a lot of polygon trees, and to a lesser extent...
  8. Ben Why did we have to scale down to 3072x3072? (multiple of 512)
  9. Convolved Feature(feature map), number of the features we want to extract(depth, number of filters ), stride, zero-padding
  10. -Deconvolution layers allow the model to use every point in the small image to “paint” a square in the larger one. -”Upsampling: use a 2*2 convolution to halve the number of feature maps→ this is one important modification in U-net: we have a large number of feature channels and allow the network to propagate context information to higher resolution layers. (This is the reason we can have a higher resolution of the output ) -White boxes represent copied feature maps from contracting path. The reason of doing this? To localize and the following layers can learn to assemble a more precise output based on these information.
  11. Basically it is a one-to-many relationship.