Sign in to view source links and access this dataset
Description
FashionFail is a dataset of 2,495 high-resolution (2400x2400 pixels) images of e-commerce products, proposed in the paper "FashionFail: Addressing Failure Cases in Fashion Object Detection and Segmentation". The dataset is split into 1,344 training images, 150 validation images, and 1,001 test images. It was authored by rizavelioglu and last updated on Hugging Face in October 2024.
Use Cases
Benchmarking object detection models based on high-resolution e-commerce product images.
Evaluating segmentation model performance on fashion items.
Training models to address failure cases in fashion image analysis.
Researching the robustness of foundation models for automated annotation tasks.
Strengths
Contains 2,495 high-resolution images at 2400x2400 pixels.
Provides a structured split of 1,344 training, 150 validation, and 1,001 test images.
Addresses a specific research problem: failure cases in fashion object detection and segmentation.
Limitations
Annotations are automatically generated by foundation models, which may introduce errors.
Column-level documentation is absent; field semantics must be inferred after download.
Row count is unknown, which may limit suitability assessment.
Provenance
Source
rizavelioglu via Hugging Face.
Collection Method
Images likely sourced from e-commerce websites.
Freshness
Last updated 2024-10-01 14:10:54; freshness should be verified.
Annotations are automatically generated; users should consult the full description on the dataset page for details.