Evotegra: Extend Dataset & Data Economy Poll

OceanDAO Round 13
#1

automobile-gfbe9da5fb_1920
automobile-gfbe9da5fb_19201920×1080 524 KB

Name of Project: Extend Dataset & Data Economy Poll

Proposal in one sentence

Extend the round 8 promotional 100k traffic dataset with annotations for licence plates and faces to enable the creation and evaluation of automated anonymization algorithms and conduct a data economy poll among members of the German AI assocation.

Description of the project and what problem is it solving:

In Ocean DAO Round 8 we commited to create a professional traffic image dataset with at least 100.000 images in ~300 categories and an estimated million objects. The dataset will soon be published on the Ocean Market with an initial data token price of 1 Ocean. The idea is to create promotional datasets that help to establish Ocean Market as a prime source of high quality data.
We propose to extend this promotional dataset next the promised 300 classes with annotations for faces and license plates. By adding these 2 classes we add significant extra value to the dataset which then can be used to train and evaluate automated anonymization algorithms which is an important requirement to assure compliance with privacy regulation such as GDPR.
Next to creating new data we propose to conduct a data economy poll among the members of the German AI association. The important questions we try to figure out are: if, which and where German AI companies would consider to buy data from sources like data markets or other intermediaries. To incentivise companies to participate in the poll we will distribute 1000 Ocean directly to the poll participants.

Category

Unleash data

Metric

Data Consume Volume

What is the final product?

  1. The final product is an extension of the promotional image dataset proposed in DAO round 8 with 2 additional classes “license plates” and “faces”.
  2. Results of the poll presented in the data consume and parameters working group.

Grant deliverables

Round 8 - Requested $10k. All Funding Deployed. Publish carry over to R11

[x] Collect 100.000 images from diverse road scenes in Germany (~2-3 month)
[x] Annotate the images in ~300 classes (~1 month)
[x] Publish the dataset in a pool for the initial price of 1 Ocean (~September 2021)
Could not be published yet. Issue reported but won’t be fixed before OceanV4.

Round 11 - Requested $3k. $3k + R8 publish carry over to R12.

[x] Extend the promotional 100k dataset with face annotations
[x] Extend the promotional 100k dataset with license plate annotations
[x] Provide 5000 annotated images

Round 12 - Requested $3k. $3k deployed from carry over. 3K + R8 publish carry over to R13.

[x] Extend the promotional 100k dataset with face annotations
[x] Extend the promotional 100k dataset with license plate annotations
[x] Provide 10000 annotated images

Round 13 - Requested $9.5k.

  1. Extend the promotional 100k dataset with face annotations
  2. Extend the promotional 100k dataset with license plate annotations
  3. Provide additional 10000 annotated images
  4. Conduct a data economy poll along with German AI association members
  5. Present results of the poll to data consume and parameters working group

R8 and R11 Deliverables

The attempt to to publish the R8 dataset on Ocean Market was unsuccessful. Due to lack of resources within Ocean Core the issue is unfortunately not going to be fixed prior to the Ocean V4 release. Therefore we put the publishing on hold until Ocean V4 release.
We finished our R11 deliverables and we will provide the annotated images in addition to our R12 grant.

Funding Requested:

9.5k USD

ROI

We see this effort as a starting point to add other promotional datasets. Our goal is to establish Ocean as a prime source for high quality data. The attractive entry price should encourage many first time users to get familiar with wallets, Ocean tokens and the Ocean Market itself. By enabling new use cases to train and test automated anonymization algorithms this will notably increase the value of the data.
The poll will help to understand the requirements of professional data consumers and providers.

Proposal Wallet Address:

0x61B15998893cC746B46C08FEdEE13a0d1b33bBa9

Team

Web: www.evotegra.de
Email: manthey@evotegra.de
Twitter: @evotegra
Country of Residence: Germany

Previous OceanDAO Grant?

Yes

#2

Hi @tmanthey ,

Thank you for applying for R13!

I am a Project-Guiding Evaluator and have assigned myself to help you.

Your proposal has been registered into the grants program, however as it stands, your R13 application is failing because we don’t know what you have completed or not in previous rounds. Please make sure all of your previous proposal updates inside Port include a Grant Deliverables Checklist with all checkboxes marked as completed.

Primarily, your previous Grant Deliverable updates are extremely hard to process and we are unable to understand how to reconcile your Round 8, 10, 11, and 12 deliverables.

Please provide a very clear and explicit explanation of your deliverables as follows, inside of this proposal.

Example:

Round 8 - Requested $10k. All Funding Deployed.

  • [x] Completed deliverable 1.
  • [x] Completed deliverable 2.
  • [x] Deliverable 3 could not be delivered. Carried over to R10.

Round 10 - Requested $15k. $10k Deployed. $5k Carry over.

  • [x] Completed deliverable 3.
  • [x] Completed deliverable 4.

Round 12 - Requested $5k. $10k deployed ($5k from carry over). All funds deployed.

  • [x] Completed deliverable 5.
  • [x] Completed deliverable 6.

Round 13 - Requested $9.5k.

  • [ ] Deliverable 7
  • [ ] Deliverable 8

Please improve how you are communicating the grant deliverable checklists (based on our directions) so that we are able to properly maintain your project & proposals to keep them in good standing.

You can find more examples of how to do this here.

All the best!
-Idiom

1 Like
#3

I updated the request. Also I noticed a typo. We did not get a grant in R10 but R11

#4

Thank you Tobias. This will help a lot.

However, I’m going to be doing some reconciliation for you, but to be clear. All of your deliverables should be marked as completed so your project can remain in good standing.

Please read the following blog post in detail so you can provide better reconciliations so that every single outcome is marked as complete.

Example:

If a deliverable is outstanding, your project will remain “In Progress” and will not be eligible to apply for another grant until every deliverable is delivered/reconciled.

All the best!

  • Idiom
#5

Ok I remember. Should be fixed with the latest update

1 Like
#6

Hi Tobias, I have now updated your previous Grant Deliverable Checklists and your project + proposals are in good standing.

At any point, you can view our Project Standing page to learn more about it and keep it up-to-date.

You can now see your latest reconciliations inside of the dahsboard.

Screenshot from 2022-01-04 13-53-58
Screenshot from 2022-01-04 13-53-581522×807 132 KB

1 Like
#7

I have made some adjustments on your behalf.

Please review and LMK if you have any issues.

1 Like
#8

Hello @tmanthey,

Please notice that we are updating the funding process for R13 and there are important things for you to know. We will be deploying an update to the Ocean Website, where you will be able to claim your grant on your own time.
Funding the smart contract may take ~2 days after Voting Ends. There will still be a funding deadline of 2 weeks to claim your grant.
You will need a web3 wallet + Airtable to complete the process. Grant requests via request.finance will be deprecated.

You can find all instructions here: https://github.com/oceanprotocol/oceandao/wiki/Request-Grant-Guidelines
You will have until January 10th to update your proposal with all required fields.

Please let us know if you experience any issues or bugs. You can also submit tickets to Github and ping us inside of ocean-dao-engineering for any help. Thank you!

#9

This is a very interesting project guys It’s very exciting for the future of the automobile! How do you collect imaging at the moment is it via use of motorway cameras etc or satellite mapping?

1 Like
#10

For traffic we collect data with in-car camera systems. These cameras are attached to embedded AI systems that evaluate images in-car real-time. Using AI only images with potentially interesting objects are collected

1 Like
#11

Wow that’s really interesting! Is it only in Germany at the moment? Will this data be widely used for all manufacturers or is it set for certain ones like Tesla/Bmw as cars develop do you plan to work to integrate with new systems as they evolve?

#12

Hey. What type of useful information can be extracted from ‘diverse road scenes’ and who will be buying the datasets?
What’s the legality of taking nonconsensual photos of people and pairing it with their licence plates?
Sorry if I missed any prior explanations about these questions.

1 Like
#13

Please have look at our dataset . There you find a link to the taxonomy describing what which ~300 objects we support.
GDPR allows data collection under the “legitimate interest” GDPR 6.1f. By not collecting irrelevant data we follow the principle of data minimisation GDPR 5.1c.
Basically the same principles that all mapping companies e.g. HERE, Google Streetview etc follow.
We are not extracting information about the specific license plate, e.g. which number. Only the location in the image where there is “a license” plate.

1 Like
#14

Hi Timothy,

Thank you for consideration and for your proposal in R13!

I love the fact that you are working towards a more accessible data set, and believe these initiatives are very important towards increasing DCV in Ocean, as we have discussed previously.

My only feedback is that I would like to see some initiatives that evolve Evotegra into perhaps a more Data-Union model, where perhaps DU participants are providing data, reviewing it, & annotating it.

As an example, a colleague worked with https://comma.ai/ to monetize vehicle data and enable everyone with a camera to monetize.

In other words, I would favor helping to expand the data economy, establish new web3 communities, and scaling the data collection/annotation pipeline rather than just increasing the number of images and type of annotations for the Evotegra dataset.

As defined in the project criteria, I believe this is one way to improve:

  • Usage of Ocean
  • Viability
  • Adding value to the community

If these changes were implemented, I believe Evotegra would increase it’s value proposition in terms of utilizing these grants towards broader improvements (rather than grants acting like a way to broker data for this 1 vehicle data set, and subsidizing the cost).

Having said that, I love your community-activeness, the fact that you’re making Evotegra data set more accessible (by lowering the cost), and continue to expand on the quality of it (i.e. license plates)

All the best!

  • Idiom
2 Likes
#15

Thanks for your suggestion. I will think about it how this could be realized.

#16

Hello @tmanthey

As part of the Project Guiding - Working Group [PG-WG], We would like to congratulate on your sustained interest in the OceanDAO and for winning the R13 grant.

We also encourage you to be part of our PG-WG Guiding group as an Attendee as well as being part of guiding the newcomers of PG-WG and OceanDAO. Say hello to us at https://discord.gg/kujC2QK9 #Project-Guiding and refer to the pinned messages to find more information regarding the WG.

Regards,

Prakash

1 Like
#17

Hi @tmanthey

Please read the updated rules on how to access funding

The portal can be found @ https://oceanprotocol.com/dao

#18

“Funding the smart contract may take ~2 days after Voting Ends.”
Question: How do you know the contract is ready without burning 100 EUR in gas fees?
Executing smart contracts on Ethereum is rather expensive (100 EUR+). Do you consider to deploy the contract on other ethereum compatible blockchains?

#19

Hi there.

The smart contract has been funded. The transaction gas fee cost for claiming your grant is about $40. We are working on finding a solution to reduce the cost of claiming a grant.

To claim your grant:

Please let me know if you have any questions!

1 Like
#20

Our results:
The images for the R8 dataset we committed including R13 have been annotated. The poll we carry over to R14.

Round 13 - Requested $9.5k. $3k deployed from carry over. 6k deployed. 3K + R8 publish carry over to R14

[x] Extend the promotional 100k dataset with face annotations
[x] Extend the promotional 100k dataset with license plate annotations
[x] Provide additional 10000 annotated images
[x] Conduct a data economy poll along with German AI association members
[x] Present results of the poll to data consume and parameters working group

1 Like