A repackaged version of the ILSVRC/ImageNet-1k dataset, transformed into a uniform 64x64 pixel resolution. The images were center-cropped to square and resampled using the Lanczos algorithm, with higher-resolution versions also available. This version was created by benjamin-paine and last updated on September 15, 2024.
Use Cases
- Train image classification models based on the standardized 64x64 pixel format.
- Benchmark model performance on a downscaled version of the ImageNet dataset.
- Develop lightweight vision models for resource-constrained environments based on the small image size.
- Conduct experiments in image compression or super-resolution using the consistent Lanczos resampling method.
Strengths
- Images are uniformly processed to a consistent 64x64 pixel resolution.
- The dataset is a complete repack of the original ILSVRC/ImageNet-1k benchmark.
- Processing used Lanczos resampling, a high-quality interpolation method.
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download.
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
Provenance
- Source
- ILSVRC/imagenet-1k
- Collection Method
- Images were center-cropped to square and rescaled to 64x64 using Lanczos resampling.
- Time Range
- null
- Freshness
- Last updated 2024-09-15 05:48:53; freshness should be verified.
- Geography
- null