What is Sensor AI?
Sensor AI is a decentralized platform that revolutionizes AI development by providing tools for data annotation, dataset sharing, and AI model training. Built on Ethereum blockchain, it uses the $SENSE token to incentivize participation, ensure transparency, and facilitate transactions within the ecosystem.
What problems does Sensor AI solve?
High costs of AI dataset annotation. Centralized control of data in traditional AI platforms. Limited access to quality datasets for small developers. Inefficient manual labeling processes.
Who can use Sensor AI?
AI Researchers & Data Scientists – Access annotated datasets and train models. Annotators – Earn $SENSE by labeling data. Businesses – Source AI-ready datasets for commercial use. Developers – Build tools on top of Sensor AI’s infrastructure.
Is Sensor AI free to use?
The basic annotation tool is free. Advanced features (marketplace purchases, model training) require $SENSE tokens.
Do I need crypto experience to use Sensor AI?
No! The platform is designed for both crypto-native and traditional AI users.
Non-crypto users can use fiat gateways (credit card purchases of $SENSE). Crypto users can directly transact with wallets.
How do I get $SENSE tokens?
Buy from supported exchanges (once listed).
Earn by completing annotation tasks.
Participate in community rewards (airdrops, referrals).
What types of data can I label?
Currently, images (with plans for text, video, and LiDAR support).
Supported annotations:
✅ Bounding boxes (rectangles)
✅ Polygons (complex shapes)
✅ Keypoints (pose estimation)
✅ Lines (lane detection, segmentation)
How does AI-assisted labeling work?
The tool integrates YOLOv5, SSD, and PoseNet to auto-suggest labels. You can:
Accept/reject AI suggestions. Manually refine labels for accuracy.
How are annotators rewarded?
Pay-per-task: Earn $SENSE for each approved annotation. Reputation system: Higher accuracy = higher earnings. Bonuses: Top annotators get extra tokens.
How is annotation quality ensured?
Multi-annotator verification: 3+ users label the same data; consensus determines correctness. Dispute system: Users can flag errors for review. Reputation penalties: Poor-quality annotators earn less.
How do I sell my dataset?
Upload your dataset (IPFS or decentralized storage). Set a price in $SENSE (fixed or auction). Define licensing (CC-BY, commercial use, exclusive rights). List it on the marketplace.
How do I buy a dataset?
Filter by domain (medical, automotive, etc.) or annotation type. Pay with $SENSE; download instantly.
Who owns the data after purchase?
Depends on the license:
Open datasets (CC-BY) – Free to use with attribution. Commercial licenses – Restricted to buyer’s use. Exclusive rights – Only the buyer can use it.
How does decentralized training work?
You submit your annotated dataset and training script. Sensor AI distributes the job across decentralized GPUs (via Golem/iExec). You pay in $SENSE based on compute time.
What frameworks are supported?
TensorFlow, PyTorch, ONNX (more coming soon).
Can I use pre-trained models?
Yes! Sensor AI offers:
Fine-tuning of existing models (YOLO, ResNet). Model zoo where users sell trained models.
What is $SENSE used for?
Payments: Buy datasets, computation, premium features. Staking: Earn rewards by locking tokens. Governance: Vote on platform upgrades.
Where can I trade $SENSE?
$SENSE will be traded on Uniswap (Ethereum)
What’s next for Sensor AI?
Mobile app for on-the-go annotation. Text/NLP labeling support. DAO governance for community-led decisions.
How can I contribute?
Develop plugins/integrations.
FAQ