documentation update

2024-09-24 22:38:34 +00:00 · 2024-09-24 22:38:34 +00:00 · 2d661b519b
--- a/.gitignore
+++ b/.gitignore
@ -17,3 +17,4 @@ PytorchWildlife.egg-info/
 *annotations.csv
 *cropped_resized*
 *log*
+MDV6b-yolov9c.pt
--- a/INSTALLATION.md
+++ b/INSTALLATION.md
@ -84,11 +84,11 @@ docker run -p 80:80 andreshdz/pytorchwildlife:1.0.2.3 python demo/gradio_demo.py
 Here is a brief example on how to perform detection and classification on a single image using `PyTorch-wildlife`:

 ```python
-import torch
+import numpy as np
 from PytorchWildlife.models import detection as pw_detection
 from PytorchWildlife.models import classification as pw_classification

-img = torch.randn((3, 1280, 1280))
+img = np.random.randn(3, 1280, 1280)

 # Detection
 detection_model = pw_detection.MegaDetectorV5() # Model weights are automatically downloaded.
@ -110,9 +110,9 @@ python image_demo.py
 # For the video demo
 python video_demo.py
 # For the gradio app
-python demo_gradio.py
+python gradio_demo.py
 ```
-The `demo_gradio.py` will launch a Gradio interface where you can:
+The `gradio_demo.py` will launch a Gradio interface where you can:
 - Perform Single Image Detection: Upload an image and set a confidence threshold to get detections.
 - Perform Batch Image Detection: Upload a zip file containing multiple images to get detections in a JSON format.
 - Perform Video Detection: Upload a video and get a processed video with detected animals. 
--- a/docs/demo/image_detection_demo.ipynb
+++ b/docs/demo/image_detection_demo.ipynb
@ -57,7 +57,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
   "id": "8f622040",
   "metadata": {},
   "outputs": [],
@ -82,7 +82,7 @@
   "outputs": [],
   "source": [
    "DEVICE = \"cuda\"  # Use \"cuda\" if GPU is available \"cpu\" if no GPU is available\n",
-    "detection_model = pw_detection.MegaDetectorV5(device=DEVICE, pretrained=True)"
+    "detection_model = pw_detection.MegaDetectorV5(device=DEVICE, pretrained=True, version=\"a\")"
   ]
  },
  {
@ -116,18 +116,15 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
   "id": "0d730b20",
   "metadata": {},
   "outputs": [],
   "source": [
    "tgt_img_path = os.path.join(demo_path, \"demo_data\", \"imgs\", \"10050028_0.JPG\")\n",
-    "img = np.array(Image.open(tgt_img_path).convert(\"RGB\"))\n",
-    "transform = pw_trans.MegaDetector_v5_Transform(target_size=detection_model.IMAGE_SIZE,\n",
-    "                                               stride=detection_model.STRIDE)\n",
-    "results = detection_model.single_image_detection(transform(img), img.shape, tgt_img_path)\n",
-    "pw_utils.save_detection_images(results, os.path.join(demo_path,\"demo_output\"))\n",
-    "print(demo_path)"
+    "results = detection_model.single_image_detection(tgt_img_path)\n",
+    "pw_utils.save_detection_images(results, os.path.join(demo_path,\"demo_output\"), overwrite=False)\n",
+    "pw_utils.save_crop_images(results, os.path.join(\".\",\"crop_output\"), overwrite=False)"
   ]
  },
  {
@ -147,16 +144,7 @@
   "outputs": [],
   "source": [
    "tgt_folder_path = os.path.join(demo_path, \"demo_data\", \"imgs\")\n",
-    "\n",
-    "dataset = pw_data.DetectionImageFolder(\n",
-    "    tgt_folder_path,\n",
-    "    transform=pw_trans.MegaDetector_v5_Transform(target_size=detection_model.IMAGE_SIZE,\n",
-    "                                                 stride=detection_model.STRIDE),\n",
-    "    extension='JPG' # Can be other extensions based on your own data, such as png. And this argument is case sensitive.\n",
-    ")\n",
-    "loader = DataLoader(dataset, batch_size=32, shuffle=False, \n",
-    "                    pin_memory=True, num_workers=0, drop_last=False)\n",
-    "results = detection_model.batch_image_detection(loader)"
+    "results = detection_model.batch_image_detection(tgt_folder_path, batch_size=16, extension=\"JPG\")"
   ]
  },
  {
@ -173,7 +161,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 8,
   "id": "f63310ab",
   "metadata": {},
   "outputs": [],
@ -192,7 +180,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 9,
   "id": "13653739",
   "metadata": {},
   "outputs": [],
@ -211,7 +199,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 10,
   "id": "b627280b",
   "metadata": {},
   "outputs": [],
@ -233,7 +221,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 11,
   "id": "eff35a43",
   "metadata": {},
   "outputs": [],
--- a/docs/demo/video_detection_demo.ipynb
+++ b/docs/demo/video_detection_demo.ipynb
@ -57,7 +57,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
   "id": "24b2cf06",
   "metadata": {},
   "outputs": [],
@ -108,27 +108,6 @@
    "classification_model = pw_classification.AI4GOpossum(device=DEVICE, pretrained=True)"
   ]
  },
-  {
-   "cell_type": "markdown",
-   "id": "fa4913d8",
-   "metadata": {},
-   "source": [
-    "## Transformations\n",
-    "Define transformations for both detection and classification. These transformations preprocess the video frames for the models."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "cc6377ee",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "trans_det = pw_trans.MegaDetector_v5_Transform(target_size=detection_model.IMAGE_SIZE,\n",
-    "                                               stride=detection_model.STRIDE)\n",
-    "trans_clf = pw_trans.Classification_Inference_Transform(target_size=224)"
-   ]
-  },
  {
   "cell_type": "markdown",
   "id": "3cd6262a",
@ -145,16 +124,18 @@
   "metadata": {},
   "outputs": [],
   "source": [
-    "box_annotator = sv.BoundingBoxAnnotator(thickness=4)\n",
+    "box_annotator = sv.BoxAnnotator(thickness=4)\n",
    "lab_annotator = sv.LabelAnnotator(text_color=sv.Color.BLACK, text_thickness=4, text_scale=2)\n",
    "\n",
    "def callback(frame: np.ndarray, index: int) -> np.ndarray:\n",
-    "    results_det = detection_model.single_image_detection(trans_det(frame), frame.shape, index)\n",
+    "    results_det = detection_model.single_image_detection(frame, img_path=index)\n",
    "    labels = []\n",
+    "\n",
    "    for xyxy in results_det[\"detections\"].xyxy:\n",
    "        cropped_image = sv.crop_image(image=frame, xyxy=xyxy)\n",
-    "        results_clf = classification_model.single_image_classification(trans_clf(Image.fromarray(cropped_image)))\n",
+    "        results_clf = classification_model.single_image_classification(cropped_image)\n",
    "        labels.append(\"{} {:.2f}\".format(results_clf[\"prediction\"], results_clf[\"confidence\"]))\n",
+    "\n",
    "    annotated_frame = lab_annotator.annotate(\n",
    "        scene=box_annotator.annotate(\n",
    "            scene=frame,\n",
@ -163,9 +144,10 @@
    "        detections=results_det[\"detections\"],\n",
    "        labels=labels,\n",
    "    )\n",
+    "    \n",
    "    return annotated_frame \n",
    "\n",
-    "pw_utils.process_video(source_path=SOURCE_VIDEO_PATH, target_path=TARGET_VIDEO_PATH, callback=callback, target_fps=5)"
+    "pw_utils.process_video(source_path=SOURCE_VIDEO_PATH, target_path=TARGET_VIDEO_PATH, callback=callback, target_fps=10)"
   ]
  },
  {
--- a/docs/user/getting_started.rst
+++ b/docs/user/getting_started.rst
@ -24,6 +24,6 @@ Tutorials

 To help you get hands-on with the package, we provide detailed tutorials that walk you through various functionalities of PytorchWildlife:

-3. **Gradio Interface**: Dive into an interactive demo using Gradio, which provides a user-friendly interface for quick demonstrations. [Link to the tutorial](./demo/gradio_demo.ipynb)
+3. **Gradio Interface**: Dive into an interactive demo using Gradio, which provides a user-friendly interface for quick demonstrations. `Link to the tutorial <./demo/gradio_demo.ipynb>`_  

 We recommend starting with the Gradio Interface tutorial for out-of-the-box inference and then progressing to the others as you become more familiar with the package.
--- a/megadetector.md
+++ b/megadetector.md
@ -39,8 +39,6 @@ We are also working on an extra-large version of MegaDetectorV6 for optimal perf
 ### 🎉 Pytorch-Wildlife ready for citation
 In addition, we have recently published a [summary paper on Pytorch-Wildlife](https://arxiv.org/abs/2405.12930). The paper has been accepted as an oral presentation at the [CV4Animals workshop](https://www.cv4animals.com/) at this year's CVPR. Please feel free to [cite us!](#cite-us)

-### 🛠️ Compatibility with CUDA 12.x
-The new version of PytorchWildlife uses the latest version of Pytorch (currently 2.3.1), which is compatible with CUDA 12.x.

 ## ✅ Feature highlights (Version 1.0.2.15)
 - [x] Added a file separation function. You can now automatically separate your files between animals and non-animals into different folders using our `detection_folder_separation` function. Please see the [Python demo file](demo/image_separation_demo.py) and [Jupyter demo](demo/image_separation_demo.ipynb)!
@ -376,3 +374,5 @@ Protected Areas Unit, Canadian Wildlife Service
 >[!IMPORTANT]
 >If you would like to be added to this list or have any questions regarding MegaDetector and Pytorch-Wildlife, please [email us](zhongqimiao@microsoft.com) or join us in our Discord channel: [![](https://img.shields.io/badge/any_text-Join_us!-blue?logo=discord&label=PytorchWildife)](https://discord.gg/TeEVxzaYtm)

+
+
--- a/requirements.txt
+++ b/requirements.txt
@ -6,4 +6,7 @@ Pillow
 supervision==0.23.0
 gradio
 ultralytics-yolov5
-chardet
+chardet
+wget
+ultralytics
+setuptools==59.5.0
--- a/setup.py
+++ b/setup.py
@ -24,7 +24,10 @@ setup(
        'gradio==4.8.0',
        'ultralytics-yolov5',
        'ultralytics',
-        'chardet'
+        'chardet',
+        'wget',
+        'ultralytics',
+        'setuptools==59.5.0'
    ],
    classifiers=[
        'Development Status :: 3 - Alpha',  
@ -35,6 +38,7 @@ setup(
        'Programming Language :: Python :: 3.7',
        'Programming Language :: Python :: 3.8',
        'Programming Language :: Python :: 3.9',
+        'Programming Language :: Python :: 3.10+',
    ],
    keywords='pytorch_wildlife, pytorch, wildlife, megadetector, conservation, animal, detection, classification',
    python_requires='>=3.8',