Project translated automatically. Log in or register, to view the original To learn neurons to recognize objects, it is necessary to collect a dataset of 7,000 photographs and disperse it by maps by object.First, it is necessary to first drop 30 photos to assess that the task is understood correctly and confirm the further work.Objects for these 30 test photos: a regiment, a door, a dryer, a tub, a carnis, a furniture door.If everything is fine, you need to collect a set of 1000 photos of each of the objects: • The field • The closet The Tumba • bed The sofa 500 photos for each of the following objects: • The door pet • The door conductor Ikea furniture box for assembly • Furniture door Required photo characteristics: the object on the photo should be in a natural environment, for example in the apartment, office, etc., not just an object on a clean background or without a background; the photo should not be watermark or any text/log; at least 5-10% of the photographs should be with clearly broken objects; the desired JPEG, GIF, BMP photo formats each image should be 30 MB or less.The main thing is that the photo on a particular object does not fall into any of the other items listed, because the model will be confused, it is also not possible that the items that are attached to the files and these (Wardrobe, entrance door, interroom door, door hand, door castle, carnis, dryer) If there the edge looks like and the type of it, it is possible.For example, in the data network, the door hand should not be visible all the door.Electronics can also not.Books, vasses, things can Specific criteria in size are not, but the main thing is not very small.

Proposals 5

Nadya Konovalenko

Winning proposal

5 0

Budget: 7000 UAH Deadline: 9 days

Good day . Ready to fulfill your task. Experience in similar work.
I will be happy to cooperate

Alina Lukash

101 1

Projects 105
Rating -
Rating 1 124

Budget: 7000 UAH Deadline: 14 days

Good day . There is experience with the collection of information, photos and much more. I will do everything according to your project. I will be happy to cooperate.

The list does not show proposals concealed by the client or freelancer with a Plus profile, as well as proposals violating rules

Current freelance projects in the category Information Gathering

Find audio

Data Processing 11 proposals 9:12

Not specified
Collect 1,000+ active Telegram investment/crypto groups

Data Processing 21 proposals 31 July

Not specified
Need consultation on professional streaming of PUBG Mobile from iPad Pro M4.

Tuition 3 proposals 30 July

22 USD
Collect active Telegram investment/crypto groups with admin and moderator contacts

Data Processing 16 proposals 30 July

Not specified
Parsing a database of decision-makers in the dating industry.

15 proposals 30 July

45 USD

Pavlo Kostenko
Kyiv, Ukraine

Projects 6
Rating -
Rating 394

Pavlo Kostenko

Nadya Konovalenko

Nadya Konovalenko

Alina Lukash

Proposals are currently absent

Current freelance projects in the category Information Gathering

Find audio

Collect 1,000+ active Telegram investment/crypto groups

Need consultation on professional streaming of PUBG Mobile from iPad Pro M4.

Collect active Telegram investment/crypto groups with admin and moderator contacts

Parsing a database of decision-makers in the dating industry.