samder03/2025-24679-image-autogluon-predictor

Model Card for Model ID

This model is an image classifier that identifies images of stop signs. It is trained with Autogluon multimodal on the ecopus/sign_identification dataset.

Model Details

Model Description

This model is an image classifier that identifies images of stop signs. It is trained with Autogluon multimodal on the ecopus/sign_identification dataset.

• Developed by: Sam Der
• Model type: AutoML (AutoGluon MultiModalPredictor with ResNet18 backbone)
• License: MIT

Uses

Direct Use

This model is intended to be used to distinguish stop signs from other street signs.

Training Details

Training Data

• dataset: ecopus/sign_identification
• splits:
  • original: 30 original images
  • augmented: 385 synthetic images

Training Procedure

• library: AutoGluon MultiModal
• presets: "medium_quality"
• backbone: timm_image → resnet18

Training Hyperparameters

• presets="medium_quality"
• hyperparameters={ "model.names": ["timm_image"], "model.timm_image.checkpoint_name": "resnet18", }

Evaluation

Testing Data, Factors & Metrics

Testing Data

ecopus/sign_identification

Metrics

• accuracy: fraction of correctly predicted labels
• F1 (weighted): harmonic mean of precision and recall, weighted by class support

Results

accuracy: 1.0000 | weighted F1: 1.0000