ResNet18 Multi-Image
Classify inputs from multiple image modalities using ResNet18
ResNet18-based model for multi-image input classification. Takes images from multiple modalities (organized in separate folders) and produces a class prediction. Requires a fine-tuned checkpoint.
When to use:
- Medical imaging with multiple scan types (e.g., MRI + CT)
- Multi-view product classification from several camera angles
- Any task combining information from multiple distinct image sources
Input:
- Finetuned Checkpoint (optional): Fine-tuned model weights
- Input Images (required): Directory containing images from multiple modalities
- Prompt (optional): Text prompt or question about the images
Output: Classification result and generation metadata
Inference Settings
No dedicated inference-time settings. Classification is determined by the loaded fine-tuned checkpoint.