Question 1

What is training dataset for Artificial Intelligence and Machine Learning?

Accepted Answer

To let your AI/ML model work in any real-life scenario it needs to be trained on a huge amount of training datasets. Based on the type of model specific type of data is used to train the model. It can be in the form of images, videos, Speech/audio or text. It can be annotated or unannotated as per the requirement.For a fully functional AI model that works on every scenario, it needs to be trained on diverse and unbiased datasets. Each AI model is continuously evolving part and so 'No Data is Enough Data' means it needs to be continuously trained on different aspects.

Question 2

What is 'Quality training data' and its importance?

Accepted Answer

Quality training data is data which is structured, unbiased and all compliance covered. To make your AI model functional for all circumstances we have to make sure that the collected data is qualified as a quality dataset.

Question 3

Datasets listed in Datastore available for Commercial use?

Accepted Answer

Yes, datasets listed on the FBAI Datastore is available for all kind of commercial use. We will share the appropriate licence with you that grants access to you for your commercial use.

Question 4

How to download the Free Sample dataset?

Accepted Answer

You can always get a free sample of all datasets listed on the datastore. You can click 'download sample' on the dataset page or you can contact us for sample data as well.

Question 5

How does FutureBeeAI follow data security compliances?

Accepted Answer

For FutureBeeAI data security matters the most. We follow all the guidelines provided by the laws of the land from where we collect the data. FutureBeeAI data infrastructure is built on the AWS S3 cloud so the data security responsibility is shared.  The IT infrastructure that AWS provides for FutureBeeAI is designed and managed in alignment with best security practices and a variety of IT security standards.

Question 6

How to get custom AI datasets which are not listed on FBAI Datastore?

Accepted Answer

Your specific requirements might be different from available datasets. You might need a mix of what we already have. Contact us to create your custom dataset. We can also facilitate the custom collection through our global network of data providers in case your requirement is really unique. Contact us to get a free consultation from an expert data manager.

Question 7

What is a custom AI dataset?

Accepted Answer

Each AI model is built for a unique purpose. It may be possible that the training dataset is not already available or that acquiring an already available dataset is very complex or highly expensive. In that case to train the AI model we opt for the custom collection. For custom collection, you might need a huge network of people with diverse attributes or technology to curate the synthetic datasets. For any custom dataset requirement, the FutureBeeAI community is the best option. Explore now.

Question 8

What if I want to collect custom data only for me? No reselling!

Accepted Answer

Totally understood. We can collect the data that is only for you. You will get all the rights of the data and no redistribution of data will be done. We can also use your custom tool in case of a requirement.

Question 9

How does FutureBeeAI promote ethical collection?

Accepted Answer

Educating data providers about the type of data we will collect, the potential use of the data, the application and insights about the data and the risk involved is a crucial part of our collection process. We take proper written consent from each data provider prior to collecting any kind of data.

Question 10

How does FutureBeeAI onboard an appropriate network for my custom collection?

Accepted Answer

We have a huge network of freelancer and vendor data providers with their specific details in our database. So, as per the specific need of your project, we reach out to them through our community channel and onboard them through a set process. We can onboard the network based on country, gender, age, race, language, accent, demographic, device, skin tone and many more things.We have access to many third-party communities as well and we also opt for various other media channels in case of some situation.

Question 11

How does FutureBeeAI educate the data provider about the potential use of their data?

Accepted Answer

We respect the data security of the provider. As the first step whenever data providers use any of our apps or platform we ask for their consent on terms of use, privacy policy and data security. Only after giving their consent is he/she allowed to continue with us. Apart from that as per the need of a specific project, we have a platform to get the NDA and consent letter signed for that particular type of data that we will collect from that particular project.  We are very limited while collecting any personal information and we only collect information which is necessary. Also, user can delete their respective account at any time and also ask for removing the information. As far as data storage and security matter we opt for the best option available for secure storage and we follow all security compliances.

Question 12

How does FutureBeeAI fairly pay the data provider for various kinds of data?

Accepted Answer

We are known for our fair pay and timely appreciation and recognition of our data providers. Pay rate totally depends on the type of data, the complexity of data, client requirements and many other factors. We always keep the data provider in our primal thoughts while deciding the payment and we try to give the best in the market rates.

Question 13

What if I want to get a small chunk of data and not the entire dataset? Is it possible?

Accepted Answer

Yes, you can always ask for a custom package. You can ask for a small part of the big dataset or maybe a mix of the various dataset in the proper percentage of the mix as per your specific need.

Question 14

I want to choose specific data from the available dataset according to various attributes like age, gender, country and accent, is it possible?

Accepted Answer

Yes, you can ask always choose the custom package according to any available attributes like age, gender etc. For example, we can help you with a dataset of males from the age group of 30-50 from India from a large dataset of facial data of south Asian people

Question 15

What is Spontaneous Dialog Speech Data and how does FutureBeeAI collect it?

Accepted Answer

Spontaneous speech data is where we collect audio of two or more people in any conversation which is unscripted and free flow. We collect two types of spontaneous speech data - General conversation and Call centre conversation. General conversation is like two or more people having a conversation on any random and regular topic. It's more like people having a phone call with friends or family.Call centre conversation is more industry, domain and scenario-specific where one person will play the role of agent and the other plays the role of customer. They are trained on the scenario prior to the conversation. We record these conversations and transcribe them to make the dataset ready to deploy.

Question 16

What is Scripted Monologue Speech Data and how does FutureBeeAI collect it?

Accepted Answer

Scripted Monologue speech collection is where the speaker will read out already scripted prompts/ sentences with various background environment as per requirement.

Question 17

How does FutureBeeAI make sense of the data?

Accepted Answer

Unstructured or biased data is of no use for your AI application. So to make sure that whatever data you are feeding to your AI model needs to be structured and unbiased. Along with collecting data from diverse environments, we focus on making it structured. So along with data, we collect metadata about the data provider and environment. We collect information like age, gender, accent, country, lighting condition, pose, time etc to make the data more enriching for your AI needs. We also do transcription, annotation and labelling to make data ready to use for your AI model.

Question 18

What is data annotation/labelling for AI training datasets?

Accepted Answer

In our quest to make the data structured we annotate and label the image, video, speech and text data. In this process, we tag each object of image or part of the audio as per the class list or object list. Using this annotated data machine learns how to identify or encounter when it gets the same class or object in real life without tagging.

Question 19

Why data annotation/ labelling needed?

Accepted Answer

Properly annotated data is a gold mine for AI applications. Before the AI model performs the function in real life it must be trained on a similar situation if not the same situation. To train the machine for such a situation we need annotated data of that particular situation. To identify people with masks, the AI model needs a huge amount of diverse annotated datasets of people wearing various kinds of masks.Annotation is a crucial part of the AI training process. You can get the help of an expert annotator team from the FBAI Community. Explore FBAI Community now.

Question 20

When I will get my AI dataset after purchase?

Accepted Answer

If you are buying the data set in the already available format you will get the access to dataset once the fund is cleared. Once we receive the payment you will instantly get access to the entire dataset. It generally takes 2-3 business days. In case you want a mix of datasets then it may take some more time for us to make the proper package as per your needs.

Question 21

How does FutureBeeAI Accept the payment on the purchase of the dataset?

Accepted Answer

We are accepting payment through various modes like bank transfer, wire transfer, PayPal, and Payoneer. We are will integrate a payment gateway soon so then you can pay through your card also.

Question 22

How will I get the Dataset after making the payment?

Accepted Answer

As soon as the payment is received you will get the custom link to our store where you will receive the final dataset as per the agreement along with all metadata.

Question 23

I want to use the dataset for Academic/ research/ non-profit functions, is there any special offer for that?

Accepted Answer

We are very much interested in an AI-enabled and human-centric world. We are always here to support academic, research and non-profit use. Feel free to reach out to us.

Question 24

Any discount offer for the batch purchase or continuous contract?

Accepted Answer

Yes, of course. If you want to purchased huge batch data we always have discounts for you. We are always looking for long-term and delighted clients. So, even for custom collection if you opt for a continuous contract we will give a special discount to you.

Question 25

Any provision for a refund?

Accepted Answer

A refund is needed when you get an unqualified dataset. We will make sure you get what you are looking for. The sample dataset is an identical part of the entire dataset. Our data manager will confirm everything before your purchase about the data you will receive, metadata and everything to avoid such a situation. So, we avoid a situation where you have to ask for it. Refund is not an option with us. In case of any functional error, you can always reach out to us. We are always available even for after-purchase assistance.

Question 26

Is there any option to generate synthetic datasets?

Accepted Answer

Yes, we have an expert tech team to generate the synthetic dataset for you. We have already listed various synthetic datasets on the store you can filter them and have a look. Feel free to reach out to us for your custom need for synthetic data.

Question 27

I already have the data, can I just get transcription/ annotation?

Accepted Answer

You are most welcome. Our expert team of annotators can help you with any kind of image, video or text annotation like bounding box, polygon, polyline, semantic or anything. In our community, we have native people for all languages so scaling your transcription project is always easy with the FutureBeeAI community.

Question 28

What if I just need raw data without transcription/annotation?

Accepted Answer

It is possible. We will share the respective quote for that and make the package for you.

Question 29

I have AI training data, how can I sell it?

Accepted Answer

If you have any kind of AI training data in the form of Image, speech, text or video you can always reach out to us via the 'sell your data' option. You can send the details and we will reach out to you for more communication.

Question 30

What tools does FutureBeeAI use?

Accepted Answer

We are lineup with our own tools for various purposes. We have Yugo which is a mobile application for all kinds of data collection needs. We have an FBAI community from which we onboard people for projects. We are coming up with our own transcription and annotation tool soon.

FAQs

Still Have Doubts?