MSCOCO-52K-FEATURES-YOLOxXCEPTION-SC2: Precomputed Image Features for Object Detection

A dataset of precomputed features for 52,000 images, likely derived from the MS COCO (Common Objects in Context) dataset. The title suggests features were extracted using YOLO and Xception neural network architectures. It is hosted on the Kaggle platform.

Use Cases

• Fine-tune an object detection model using precomputed features (inferred from domain, verify after download)
• Benchmark feature extraction methods for the COCO dataset (inferred from domain, verify after download)
• Perform transfer learning for downstream vision tasks (inferred from domain, verify after download)

Strengths

• Published on Kaggle, a major platform for data science resources.
• Title indicates a substantial scale of 52,000 feature vectors.

Limitations

• Metadata is minimal; actual content requires verification after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect the geographic, temporal, or categorical biases inherent to the original MS COCO source.

Provenance

Source

kaggle