Activity Feed

Profile pic

Johnny C. added photos to project Fungi Barbarian

Medium 18d6f30d 25ad 4e57 b999 92cd9ed616de

Fungi Barbarian

Demo:

- Blog: http://fungai.org
- ReactJS UI POC: https://fungai-react-ui.herokuapp.com/fungpredict
- Fake json-server POC: https://fungai-json-server-heroku.herokuapp.com/

Fungi Barbarian is a wild mushroom ID machine. It's goal is to assist a human identifying wild mushrooms: given some sample wild mushroom images, Fungi Barbarian will try and identify the mushroom for you. It does not require internet access and is able to perform the inferences offline. This project is originally motivated and inspired by a friends Facebook post: "hey do you know what this mushroom is?".

Core Technology Stack (for offline training and inference):

- Model Training: Intel Xeon Phi Cluster: to train the image classification model - leveraging Transfer Learning on Google Inception V3 model.
- Prediction / Inference: Movidius Neural Compute Stick (NCS), loaded with a pre-trained Tensorflow image classification model specialises in wild mushroom images
- Hardware: Raspberry Pi (e.g. MyCroft AI), or whatever suitable that is able to link up camera (for image capture), NCS, and portable power via USB connection. The entire inference solution is offline and does not require internet access.

Secondary Technology Stack (for online inference):

- Full-stack web application to enable inferences via a web application (in case user prefer to do inferences in batches after their wild-mushroom ID trips). Frontend: ReactJS single page web app. API calls with backends (e.g. tensorflow model exposed via Tensorflow, ImageNet image URLs to grab images for training, validation, and testing (inferences),

Training Data Repository:

- ImageNet: http://image-net.org/download-imageurls
- Wikimedia: to store training data (images), with free / open / permissive licence.
- or whatever repositories suitable for storing open data in suitable format (e.g. images, video, sensor data, etc.)

Documentations

- Blog Posts
- Interactive full-stack web application for educational purposes.

Vision

- Free everything: free data (Wikimedia?), free software (GitHub repository), free instructions (via documentations). It is important anyone can replicate / create their own solution / be inspired from this project, without the hindrance of copy right issues. The licence for this project will be highly permissible (e.g. MIT).
- Collaborate with local wild mushroom ID (human) experts on building up an reliable / accurate / trustworthy data repository on Wild Mushroom. Data should contains features (e.g. images, videos, sensor data, etc.) and labels (annotated / identified by a human expert). i.e. Whatever it is required for Supervised Training in Machine Learning terms.
- sustainable ecosystem with strong governance (something to think about).
- potential ongoing sponsorship opportunities from local nature reserve / parks / communities (?)

Foreseeable challenges and strategies:

- As there are literally thousands (if not more) species of wild mushrooms out there, it could be a challenge to collect sufficient training data. To make this project practical, it may be wise to start off with just a small group of well known categories (e.g ImageNet categories: "Earthstar", "Agaric", etc. ) - say pick 10 categories. Train the model very well against these specialised categories. Only when we have a reasonable good proof of concept we then move on to repeat process for other mushroom categories / species.
- Most images found on the web have copyright protection on them, and hence may not be suitable for consumption for this project. To work around this, try focuses on free and open images only (e.g. ImageNet). If we are going to generate our own data, make sure we keep our data free and open - to avoid copyright hindrance.
- For the purpose of initial proof of concept, it might be wise to pick a ImageNet category at an appropriate level to ensure images are sufficient "different" to the human eye for good demo purposes (The sub categories underneath the Agaric category is pretty good for this). In future if we wish to go deeper (e.g. to differentiate the different types of Earthstars), we may need to perform a more specialized form of transfer learning - to make the differentiation more sensitive. Think CT Scan images but Fungus images.
- ImageNet ImageNet ImageNet.

Profile pic

Johnny C. added photos to project Fungi Barbarian

Medium bd23fe00 6573 488f 9c2b 51d36b597a22

Fungi Barbarian

Demo:

- Blog: http://fungai.org
- ReactJS UI POC: https://fungai-react-ui.herokuapp.com/fungpredict
- Fake json-server POC: https://fungai-json-server-heroku.herokuapp.com/

Fungi Barbarian is a wild mushroom ID machine. It's goal is to assist a human identifying wild mushrooms: given some sample wild mushroom images, Fungi Barbarian will try and identify the mushroom for you. It does not require internet access and is able to perform the inferences offline. This project is originally motivated and inspired by a friends Facebook post: "hey do you know what this mushroom is?".

Core Technology Stack (for offline training and inference):

- Model Training: Intel Xeon Phi Cluster: to train the image classification model - leveraging Transfer Learning on Google Inception V3 model.
- Prediction / Inference: Movidius Neural Compute Stick (NCS), loaded with a pre-trained Tensorflow image classification model specialises in wild mushroom images
- Hardware: Raspberry Pi (e.g. MyCroft AI), or whatever suitable that is able to link up camera (for image capture), NCS, and portable power via USB connection. The entire inference solution is offline and does not require internet access.

Secondary Technology Stack (for online inference):

- Full-stack web application to enable inferences via a web application (in case user prefer to do inferences in batches after their wild-mushroom ID trips). Frontend: ReactJS single page web app. API calls with backends (e.g. tensorflow model exposed via Tensorflow, ImageNet image URLs to grab images for training, validation, and testing (inferences),

Training Data Repository:

- ImageNet: http://image-net.org/download-imageurls
- Wikimedia: to store training data (images), with free / open / permissive licence.
- or whatever repositories suitable for storing open data in suitable format (e.g. images, video, sensor data, etc.)

Documentations

- Blog Posts
- Interactive full-stack web application for educational purposes.

Vision

- Free everything: free data (Wikimedia?), free software (GitHub repository), free instructions (via documentations). It is important anyone can replicate / create their own solution / be inspired from this project, without the hindrance of copy right issues. The licence for this project will be highly permissible (e.g. MIT).
- Collaborate with local wild mushroom ID (human) experts on building up an reliable / accurate / trustworthy data repository on Wild Mushroom. Data should contains features (e.g. images, videos, sensor data, etc.) and labels (annotated / identified by a human expert). i.e. Whatever it is required for Supervised Training in Machine Learning terms.
- sustainable ecosystem with strong governance (something to think about).
- potential ongoing sponsorship opportunities from local nature reserve / parks / communities (?)

Foreseeable challenges and strategies:

- As there are literally thousands (if not more) species of wild mushrooms out there, it could be a challenge to collect sufficient training data. To make this project practical, it may be wise to start off with just a small group of well known categories (e.g ImageNet categories: "Earthstar", "Agaric", etc. ) - say pick 10 categories. Train the model very well against these specialised categories. Only when we have a reasonable good proof of concept we then move on to repeat process for other mushroom categories / species.
- Most images found on the web have copyright protection on them, and hence may not be suitable for consumption for this project. To work around this, try focuses on free and open images only (e.g. ImageNet). If we are going to generate our own data, make sure we keep our data free and open - to avoid copyright hindrance.
- For the purpose of initial proof of concept, it might be wise to pick a ImageNet category at an appropriate level to ensure images are sufficient "different" to the human eye for good demo purposes (The sub categories underneath the Agaric category is pretty good for this). In future if we wish to go deeper (e.g. to differentiate the different types of Earthstars), we may need to perform a more specialized form of transfer learning - to make the differentiation more sensitive. Think CT Scan images but Fungus images.
- ImageNet ImageNet ImageNet.

Profile pic

Johnny C. added photos to project Fungi Barbarian

Medium a00db498 eb72 4bf2 8336 25a79431e111

Fungi Barbarian

Demo:

- Blog: http://fungai.org
- ReactJS UI POC: https://fungai-react-ui.herokuapp.com/fungpredict
- Fake json-server POC: https://fungai-json-server-heroku.herokuapp.com/

Fungi Barbarian is a wild mushroom ID machine. It's goal is to assist a human identifying wild mushrooms: given some sample wild mushroom images, Fungi Barbarian will try and identify the mushroom for you. It does not require internet access and is able to perform the inferences offline. This project is originally motivated and inspired by a friends Facebook post: "hey do you know what this mushroom is?".

Core Technology Stack (for offline training and inference):

- Model Training: Intel Xeon Phi Cluster: to train the image classification model - leveraging Transfer Learning on Google Inception V3 model.
- Prediction / Inference: Movidius Neural Compute Stick (NCS), loaded with a pre-trained Tensorflow image classification model specialises in wild mushroom images
- Hardware: Raspberry Pi (e.g. MyCroft AI), or whatever suitable that is able to link up camera (for image capture), NCS, and portable power via USB connection. The entire inference solution is offline and does not require internet access.

Secondary Technology Stack (for online inference):

- Full-stack web application to enable inferences via a web application (in case user prefer to do inferences in batches after their wild-mushroom ID trips). Frontend: ReactJS single page web app. API calls with backends (e.g. tensorflow model exposed via Tensorflow, ImageNet image URLs to grab images for training, validation, and testing (inferences),

Training Data Repository:

- ImageNet: http://image-net.org/download-imageurls
- Wikimedia: to store training data (images), with free / open / permissive licence.
- or whatever repositories suitable for storing open data in suitable format (e.g. images, video, sensor data, etc.)

Documentations

- Blog Posts
- Interactive full-stack web application for educational purposes.

Vision

- Free everything: free data (Wikimedia?), free software (GitHub repository), free instructions (via documentations). It is important anyone can replicate / create their own solution / be inspired from this project, without the hindrance of copy right issues. The licence for this project will be highly permissible (e.g. MIT).
- Collaborate with local wild mushroom ID (human) experts on building up an reliable / accurate / trustworthy data repository on Wild Mushroom. Data should contains features (e.g. images, videos, sensor data, etc.) and labels (annotated / identified by a human expert). i.e. Whatever it is required for Supervised Training in Machine Learning terms.
- sustainable ecosystem with strong governance (something to think about).
- potential ongoing sponsorship opportunities from local nature reserve / parks / communities (?)

Foreseeable challenges and strategies:

- As there are literally thousands (if not more) species of wild mushrooms out there, it could be a challenge to collect sufficient training data. To make this project practical, it may be wise to start off with just a small group of well known categories (e.g ImageNet categories: "Earthstar", "Agaric", etc. ) - say pick 10 categories. Train the model very well against these specialised categories. Only when we have a reasonable good proof of concept we then move on to repeat process for other mushroom categories / species.
- Most images found on the web have copyright protection on them, and hence may not be suitable for consumption for this project. To work around this, try focuses on free and open images only (e.g. ImageNet). If we are going to generate our own data, make sure we keep our data free and open - to avoid copyright hindrance.
- For the purpose of initial proof of concept, it might be wise to pick a ImageNet category at an appropriate level to ensure images are sufficient "different" to the human eye for good demo purposes (The sub categories underneath the Agaric category is pretty good for this). In future if we wish to go deeper (e.g. to differentiate the different types of Earthstars), we may need to perform a more specialized form of transfer learning - to make the differentiation more sensitive. Think CT Scan images but Fungus images.
- ImageNet ImageNet ImageNet.

Profile pic

Johnny C. added photos to project Fungi Barbarian

Medium f9457979 1237 4255 b29d 62a8649055ed

Fungi Barbarian

Demo:

- Blog: http://fungai.org
- ReactJS UI POC: https://fungai-react-ui.herokuapp.com/fungpredict
- Fake json-server POC: https://fungai-json-server-heroku.herokuapp.com/

Fungi Barbarian is a wild mushroom ID machine. It's goal is to assist a human identifying wild mushrooms: given some sample wild mushroom images, Fungi Barbarian will try and identify the mushroom for you. It does not require internet access and is able to perform the inferences offline. This project is originally motivated and inspired by a friends Facebook post: "hey do you know what this mushroom is?".

Core Technology Stack (for offline training and inference):

- Model Training: Intel Xeon Phi Cluster: to train the image classification model - leveraging Transfer Learning on Google Inception V3 model.
- Prediction / Inference: Movidius Neural Compute Stick (NCS), loaded with a pre-trained Tensorflow image classification model specialises in wild mushroom images
- Hardware: Raspberry Pi (e.g. MyCroft AI), or whatever suitable that is able to link up camera (for image capture), NCS, and portable power via USB connection. The entire inference solution is offline and does not require internet access.

Secondary Technology Stack (for online inference):

- Full-stack web application to enable inferences via a web application (in case user prefer to do inferences in batches after their wild-mushroom ID trips). Frontend: ReactJS single page web app. API calls with backends (e.g. tensorflow model exposed via Tensorflow, ImageNet image URLs to grab images for training, validation, and testing (inferences),

Training Data Repository:

- ImageNet: http://image-net.org/download-imageurls
- Wikimedia: to store training data (images), with free / open / permissive licence.
- or whatever repositories suitable for storing open data in suitable format (e.g. images, video, sensor data, etc.)

Documentations

- Blog Posts
- Interactive full-stack web application for educational purposes.

Vision

- Free everything: free data (Wikimedia?), free software (GitHub repository), free instructions (via documentations). It is important anyone can replicate / create their own solution / be inspired from this project, without the hindrance of copy right issues. The licence for this project will be highly permissible (e.g. MIT).
- Collaborate with local wild mushroom ID (human) experts on building up an reliable / accurate / trustworthy data repository on Wild Mushroom. Data should contains features (e.g. images, videos, sensor data, etc.) and labels (annotated / identified by a human expert). i.e. Whatever it is required for Supervised Training in Machine Learning terms.
- sustainable ecosystem with strong governance (something to think about).
- potential ongoing sponsorship opportunities from local nature reserve / parks / communities (?)

Foreseeable challenges and strategies:

- As there are literally thousands (if not more) species of wild mushrooms out there, it could be a challenge to collect sufficient training data. To make this project practical, it may be wise to start off with just a small group of well known categories (e.g ImageNet categories: "Earthstar", "Agaric", etc. ) - say pick 10 categories. Train the model very well against these specialised categories. Only when we have a reasonable good proof of concept we then move on to repeat process for other mushroom categories / species.
- Most images found on the web have copyright protection on them, and hence may not be suitable for consumption for this project. To work around this, try focuses on free and open images only (e.g. ImageNet). If we are going to generate our own data, make sure we keep our data free and open - to avoid copyright hindrance.
- For the purpose of initial proof of concept, it might be wise to pick a ImageNet category at an appropriate level to ensure images are sufficient "different" to the human eye for good demo purposes (The sub categories underneath the Agaric category is pretty good for this). In future if we wish to go deeper (e.g. to differentiate the different types of Earthstars), we may need to perform a more specialized form of transfer learning - to make the differentiation more sensitive. Think CT Scan images but Fungus images.
- ImageNet ImageNet ImageNet.

Medium profile pic

Johnny C. created project Fungi Barbarian

Medium 995fcf80 a03e 4963 96ce a44c9c6c23a7

Fungi Barbarian

Demo:

Fungi Barbarian is a wild mushroom ID machine. It's goal is to assist a human identifying wild mushrooms: given some sample wild mushroom images, Fungi Barbarian will try and identify the mushroom for you. It does not require internet access and is able to perform the inferences offline. This project is originally motivated and inspired by a friends Facebook post: "hey do you know what this mushroom is?".

Core Technology Stack (for offline training and inference):

  • Model Training: Intel Xeon Phi Cluster: to train the image classification model - leveraging Transfer Learning on Google Inception V3 model.
  • Prediction / Inference: Movidius Neural Compute Stick (NCS), loaded with a pre-trained Tensorflow image classification model specialises in wild mushroom images
  • Hardware: Raspberry Pi (e.g. MyCroft AI), or whatever suitable that is able to link up camera (for image capture), NCS, and portable power via USB connection. The entire inference solution is offline and does not require internet access.

Secondary Technology Stack (for online inference):

  • Full-stack web application to enable inferences via a web application (in case user prefer to do inferences in batches after their wild-mushroom ID trips). Frontend: ReactJS single page web app. API calls with backends (e.g. tensorflow model exposed via Tensorflow, ImageNet image URLs to grab images for training, validation, and testing (inferences),

Training Data Repository:

  • ImageNet: http://image-net.org/download-imageurls
  • Wikimedia: to store training data (images), with free / open / permissive licence.
  • or whatever repositories suitable for storing open data in suitable format (e.g. images, video, sensor data, etc.)

Documentations

  • Blog Posts
  • Interactive full-stack web application for educational purposes.

Vision

  • Free everything: free data (Wikimedia?), free software (GitHub repository), free instructions (via documentations). It is important anyone can replicate / create their own solution / be inspired from this project, without the hindrance of copy right issues. The licence for this project will be highly permissible (e.g. MIT).
  • Collaborate with local wild mushroom ID (human) experts on building up an reliable / accurate / trustworthy data repository on Wild Mushroom. Data should contains features (e.g. images, videos, sensor data, etc.) and labels (annotated / identified by a human expert). i.e. Whatever it is required for Supervised Training in Machine Learning terms.
  • sustainable ecosystem with strong governance (something to think about).
  • potential ongoing sponsorship opportunities from local nature reserve / parks / communities (?)

Foreseeable challenges and strategies:

  • As there are literally thousands (if not more) species of wild mushrooms out there, it could be a challenge to collect sufficient training data. To make this project practical, it may be wise to start off with just a small group of well known categories (e.g ImageNet categories: "Earthstar", "Agaric", etc. ) - say pick 10 categories. Train the model very well against these specialised categories. Only when we have a reasonable good proof of concept we then move on to repeat process for other mushroom categories / species.
  • Most images found on the web have copyright protection on them, and hence may not be suitable for consumption for this project. To work around this, try focuses on free and open images only (e.g. ImageNet). If we are going to generate our own data, make sure we keep our data free and open - to avoid copyright hindrance.
  • For the purpose of initial proof of concept, it might be wise to pick a ImageNet category at an appropriate level to ensure images are sufficient "different" to the human eye for good demo purposes (The sub categories underneath the Agaric category is pretty good for this). In future if we wish to go deeper (e.g. to differentiate the different types of Earthstars), we may need to perform a more specialized form of transfer learning - to make the differentiation more sensitive. Think CT Scan images but Fungus images.
  • ImageNet ImageNet ImageNet.

See More

About

Professionally I’ve been a technologist for a large global investment bank, an analytics consultant for a major UK commercial and private bank, a full-stack developer for a wellness startup, and briefly an engineering intern for an airline. In my spare time I am the creator and author of Mathalope.co.uk (a tech blog visited by 120,000+ students and professionals from 180+ countries so far), volunteer developer of the Friends of Russia Dock Woodland Website, open source software contributor, hackathon competitor, and part of the Intel® Software Innovator Program. I am currently working on becoming a better machine learning engineer and educator. Currently building fungAI.org.

Featured Projects

See All

Thumb 995fcf80 a03e 4963 96ce a44c9c6c23a7
  • Collaborators 0
  • Followers 0

Follow

Fungi Barbarian

Thumb profile pic Johnny Chan

Created: 11/09/2017

Fungi Barbarian uses deep learning to help identifying wild mushrooms from images offline.

Golden bfff1346 49d7 4ee1 b07c ffde7c78d65b
  • Members 343

DeepLearning

Practical Deep Learning and Machine learning Projects

Bigger kayatt
Innovator
  • Projects 5
  • Followers 64

Pedro Monteiro Kayatt

Master in Computer Engineering by USP and MSc. in Computer Engineering by Politecnico di Milano. Winner of some awards as TNW - Hack Battle Latin America, Top10 projects with Badaboom in Intel Perceptual Challenge Brazil in 2013 and was the first Brazilian "Intel Innovator" in 2014. He founded in 2012 the Naked Monkey Games, that has pivoted to VRMonkey in 2015.

Av. Prof. Lineu Prestes, 2242, Cietec Sala 205 - Vila Universitaria, São Paulo - SP, 05508-000, Brazil

Bigger pooja
Innovator
  • Projects 5
  • Followers 127

Pooja Baraskar

Pooja Baraskar is an Intel Innovator and Microsoft MVP. She is a Microsoft Certified Professional and C# specialist. She enjoys developing games rather than playing. She has develped many apps and games for Windows Store and Tizen Store. Her current interest is Internet of Things, hence won Microsoft Internet of Things competition and has been awarded at Azure Conference 2015 Pune also all her articles at CodeProject are prize winning. Recently she has been awarded as the Most Valuable Blogger by DZone. Know more about Pooja from her blog http://poojabaraskar.com/

Betul, Madhya Pradesh, India

Bigger screenshot www.facebook.com 2016 03 31 16 28 35
  • Projects 0
  • Followers 67

Wendy Boswell

Program Manager, Intel Software Innovators

Portland, OR, USA

Bigger peter ma
Innovator
  • Projects 8
  • Followers 90

Peter Ma

Peter Ma has been a software developer for more than 12 years. He was involved in many startups and projects that pushed the envelope of innovation. His accomplishments include speaking about mobile apps at TEDGlobal 2010. He also won the 2015 AT&T Developer Summit grand prize. In 2016, Peter gave Secretary of State John Kerry a demo of Anti-Snoozer, a drowsy detection system that leverages both Intel® EdisonTM board and Intel RealSense technology. He is part of Intel Software Innovator. Innovator.

San Francisco, CA, USA

Bigger mdp square loghi big
Black Belt
  • Projects 4
  • Followers 75

Marco Dal Pino

Microsoft MVP, Intel Black Belt Software Developer, Intel Software Innovator, Nokia Developer Champion, , Freelance Consultant, Mobile and E

55049 Torre del Lago, Province of Lucca, Italy

Bigger mjvkm9o8
Black Belt
  • Projects 4
  • Followers 67

Martin Foertsch

Intel Software Innovator and Black Belt, Oracle JavaOne Rock Star, Computer Scientist, Speaker, Author for ParrotsOnJava.com

Beta-Straße 13A, 85774 Unterföhring, Germany

See More

Thumb 6eeaf9ae 106f 4d25 9284 2ec03ee06ddd
  • Collaborators 2
  • Followers 25

Follow

b08 the Robot

Thumb steve intel mesh dev profile Steve Favis

Created: 08/16/2016

b08 the Robot is in pre-seed stage and we are recruiting developers of all skills sets and of all...

Thumb ba989cc0 818e 4b9d b1a2 63f240c1519f
  • Collaborators 0
  • Followers 2

Follow

Smart Book

Thumb 10437655 1004398386254560 343153564771573278 n zayen chagra

Created: 12/05/2016

Smart Book is one of the academic projects we are mentoring at ESPRIT Mobile. It consists on a 3D...

Medium real sense
Featured
  • Followers 1864

Intel RealSense™

Natural interaction, immersive, collaboration, gaming and learning, 3D scanning. Integrate tracki...

Medium android
Featured
  • Followers 1888

Android

Intel is inside more and more Android devices, and we have tools and resources to make your app d...

Medium big data
Featured
  • Followers 1955

Modern Code

Drive faster breakthroughs through faster code: Get more results on your hardware today and carry...

Medium networking
Featured
  • Followers 1781

Networking

Software-Defined Networking (SDN) and Network Functions Virtualization (NFV) are transforming the...

Medium achievement unlocked logo
Featured
  • Followers 1712

Game Development

Upgrade your skills as a game developer, share your game projects, and connect with other develop...

Bigger img 4504
  • Projects 0
  • Followers 2

Stephanie Essin (Intel)

I received a Masters of Science in Multimedia and Journalism at the University of Oregon in 2016. I have been with Intel since 2015, and work as the associate video producer for Intel Software TV.

United States

Bigger id kaart
  • Projects 0
  • Followers 0

Suniel Rampadarath

Born on 20th july 1971. 43 years old and 15 years experience in the field of ICT,Sms, MIS, Digital Security, differentiate on DM -field (Digital Marketing)

Paramaribo, Suriname

Bigger photo
  • Projects 0
  • Followers 1

Kaibo Liu

PhD student at Oregon State University

Corvallis, OR, USA

See More