Image Segmentation
Transformers
Safetensors
semantic-segmentation
drone
rgb
thermal
infrared
dinov3
aerial
Instructions to use markus-42/SegFly-Firefly with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use markus-42/SegFly-Firefly with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-segmentation", model="markus-42/SegFly-Firefly")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("markus-42/SegFly-Firefly", dtype="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "architectures": ["FireflyForSemanticSegmentationThermal"], | |
| "model_type": "firefly-thermal", | |
| "num_labels": 15, | |
| "backbone": "dinov3_vitb16", | |
| "image_size": 640, | |
| "embedding_dim": 256, | |
| "backbone_embed_dim": 768, | |
| "patch_size": 16, | |
| "num_layers": 12, | |
| "rein_token_length": 100, | |
| "feature_layers": [2, 5, 8, 11], | |
| "dropout_ratio": 0.1, | |
| "semantic_loss_ignore_index": 255, | |
| "id2label": { | |
| "0": "road", | |
| "1": "walkway", | |
| "2": "dirt", | |
| "3": "gravel", | |
| "4": "grass", | |
| "5": "vegetation", | |
| "6": "tree", | |
| "7": "ground_obstacles", | |
| "8": "vehicle", | |
| "9": "water", | |
| "10": "building", | |
| "11": "roof", | |
| "12": "parking_lot", | |
| "13": "constructions", | |
| "14": "truck" | |
| }, | |
| "label2id": { | |
| "road": 0, | |
| "walkway": 1, | |
| "dirt": 2, | |
| "gravel": 3, | |
| "grass": 4, | |
| "vegetation": 5, | |
| "tree": 6, | |
| "ground_obstacles": 7, | |
| "vehicle": 8, | |
| "water": 9, | |
| "building": 10, | |
| "roof": 11, | |
| "parking_lot": 12, | |
| "constructions": 13, | |
| "truck": 14 | |
| }, | |
| "transformers_version": "4.37.2" | |
| } | |