Use Cases

So far, opencrowd has been used in:

• object detection across a variety of topics, from aerial imagery to chandeliers
• during a critical mistake involving large-scale data, opencrowd was rapidly deployed to differentiate between impacted and non-impacted states. The time and accuracy was superior to the rapid sentiment analysis models deployed
• classification across 140 labels
• classification models using opencrowd for further self-evaluation in uncertain conditions
• model comparison testing: which model performed better to human standards?
• Select-k data acquisition: Which k of n in the dataset do humans prefer?
• Active content moderation and reinforcement learning