Initial commit

README.md

@@ -1,3 +1,4 @@
----
-license: mit
----
+# Usage:
+```bash
+python detector.py --image "path_to_image.png|.jpg|.jpeg"
+```

config.json

@@ -0,0 +1,63 @@
+{
+  "_name_or_path": "../models/checkpoint-4266",
+  "_num_labels": 2,
+  "architectures": [
+    "SwinForImageClassification"
+  ],
+  "attention_probs_dropout_prob": 0.0,
+  "depths": [
+    2,
+    2,
+    18,
+    2
+  ],
+  "drop_path_rate": 0.1,
+  "embed_dim": 128,
+  "encoder_stride": 32,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.0,
+  "hidden_size": 1024,
+  "id2label": {
+    "0": "artificial",
+    "1": "human"
+  },
+  "image_size": 224,
+  "initializer_range": 0.02,
+  "label2id": {
+    "artificial": 0,
+    "human": 1
+  },
+  "layer_norm_eps": 1e-05,
+  "mlp_ratio": 4.0,
+  "model_type": "swin",
+  "num_channels": 3,
+  "num_heads": [
+    4,
+    8,
+    16,
+    32
+  ],
+  "num_layers": 4,
+  "out_features": [
+    "stage4"
+  ],
+  "out_indices": [
+    4
+  ],
+  "padding": "max_length",
+  "patch_size": 4,
+  "path_norm": true,
+  "problem_type": "single_label_classification",
+  "qkv_bias": true,
+  "stage_names": [
+    "stem",
+    "stage1",
+    "stage2",
+    "stage3",
+    "stage4"
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.46.2",
+  "use_absolute_embeddings": false,
+  "window_size": 7
+}

detector.py

@@ -0,0 +1,32 @@
+import argparse
+import os
+import warnings
+from transformers import SwinForImageClassification, AutoImageProcessor
+from PIL import Image
+import torch
+
+warnings.filterwarnings("ignore")
+os.environ["TRANSFORMERS_VERBOSITY"] = "error"
+
+def classify_image(image_path):
+    model = SwinForImageClassification.from_pretrained("./")
+    processor = AutoImageProcessor.from_pretrained("./", use_fast=False)
+
+    image = Image.open(image_path).convert("RGB")
+    inputs = processor(images=image, return_tensors="pt")
+
+    with torch.no_grad():
+        outputs = model(**inputs)
+        logits = outputs.logits
+        predicted_class = torch.argmax(logits, dim=1).item()
+
+    label = model.config.id2label[predicted_class]
+    return "Real" if label == "human" else "Fake"
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Detecta se uma imagem é real ou gerada por IA.")
+    parser.add_argument("--image", type=str, required=True, help="Caminho para a imagem a ser analisada.")
+    
+    args = parser.parse_args()
+    result = classify_image(args.image)
+    print(result)

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7076a3794f258075c499156c3f44d0d2528426694e74993549270d5d362b413b
+size 347498816

preprocessor_config.json

@@ -0,0 +1,22 @@
+{
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "ViTImageProcessor",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "height": 224,
+    "width": 224
+  }
+}