State-of-the-Art Models Timeline with Multiple Metrics
This visualization shows the evolution of model performance over time across different metrics. Use the dropdown to switch between metrics. The line represents the cumulative best (SOTA) score achieved up to each point in time.
Total models in dataset: 30
Date range: 2012-01-15 to 2022-11-15
Select Metric
About this visualization:
- Metric Selector: Use the dropdown in the upper right to switch between different performance metrics
- Cyan line: Cumulative best (SOTA) score over time for the selected metric
- Cyan dots: Models that achieved a new SOTA when released
- Gray dots: Other models that didn't beat the existing SOTA
- Hover over points: See model names, release dates, and metric values
Available Metrics:
- Accuracy: Top-1 accuracy on ImageNet (%)
- Top5 Accuracy: Top-5 accuracy on ImageNet (%)
- Parameters (Millions): Model size in millions of parameters
- FLOPs (Billions): Computational cost in billions of operations
- Inference Time (ms): Time to process a single image