docs: dedicated Getting Started page (own the onboarding query) + HowTo schema

docs/getting-started.md

@@ -0,0 +1,98 @@
+---
+title: "Getting Started with PyTorch-Wildlife: Detection and Classification"
+description: "Getting started with PyTorch-Wildlife: install the framework, run your first MegaDetector detection, add a species classifier, and scale up to batch."
+slug: getting-started
+tags:
+  - getting started
+  - PyTorch-Wildlife
+  - wildlife AI framework
+  - quickstart
+  - detection and classification
+---
+
+# Getting Started with PyTorch-Wildlife
+
+New to PyTorch-Wildlife? This page is the short path from a fresh install to real output. It covers four steps: install the framework, run your first detection, add a species classifier on top, and scale up to whole folders. Each step links to the deeper reference page when you want more detail.
+
+If you only need the minimal snippet, the [Overview](index.md#quick-start) has it. If you want the guided version, read on.
+
+## Is PyTorch-Wildlife right for my project?
+
+PyTorch-Wildlife is the unified open-source framework from the Microsoft AI for Good Lab for wildlife monitoring. It bundles detection models (including MegaDetector), species classifiers, bioacoustic models, and the data utilities that tie them together, so you can run a full detect-then-classify pipeline from one package rather than stitching tools together yourself.
+
+Reach for it when you want to:
+
+- Run MegaDetector and a species classifier through one consistent Python API.
+- Process camera-trap images one at a time or thousands in a batch.
+- Build on a shared foundation instead of maintaining your own inference harness.
+
+If you only need the detector and nothing else, the dedicated [MegaDetector](https://microsoft.github.io/MegaDetector/) site is a narrower starting point. For field hardware, see [SPARROW](https://microsoft.github.io/SPARROW/); for audio, see MegaDetector-Acoustic.
+
+## Step 1: Install the framework
+
+Install from PyPI:
+
+```bash
+pip install PytorchWildlife
+```
+
+A clean conda environment keeps dependencies tidy:
+
+```bash
+conda create -n pytorch-wildlife python=3.10 -y
+conda activate pytorch-wildlife
+pip install PytorchWildlife
+```
+
+A CUDA-capable NVIDIA GPU gives a large speedup but is optional. GPU-enabled PyTorch, a Docker image, and platform notes for Windows, macOS, and Ubuntu are on the [Installation](installation.md) page. To skip installation entirely while you evaluate, try the [Hugging Face demo](https://huggingface.co/spaces/ai-for-good-lab/pytorch-wildlife) in a browser or the [Google Colab notebook](https://colab.research.google.com/drive/1rjqHrTMzEHkMualr4vB55dQWCsCKMNXi?usp=sharing) on a free cloud GPU.
+
+## Step 2: Run your first detection
+
+Load MegaDetectorV6 and detect one image. Weights download on first use:
+
+```python
+from PytorchWildlife.models import detection as pw_detection
+
+detector = pw_detection.MegaDetectorV6()
+results = detector.single_image_detection("path/to/image.jpg")
+```
+
+That is the smallest useful run. For annotated images, saved crops, and JSON exports (including a Timelapse-compatible file), the [Inference Examples](inference-examples.md) page has runnable scripts.
+
+## Step 3: Add species classification
+
+MegaDetector finds animals but does not name the species. To get species, pass each detection crop to a classifier. The two stages share the same API, so chaining them is direct:
+
+```python
+from PytorchWildlife.models import detection as pw_detection
+from PytorchWildlife.models import classification as pw_classification
+
+detector = pw_detection.MegaDetectorV6()
+classifier = pw_classification.AI4GAmazonRainforest()
+
+detection_result = detector.single_image_detection("image.jpg")
+classification_result = classifier.single_image_classification("image.jpg")
+```
+
+The framework ships several classifiers (Amazon Rainforest, Snapshot Serengeti, Opossum, Deepfaune, DFNE). The [API overview](api.md#detect-then-classify) shows the full crop-and-classify loop, and the [Model Zoo](model_zoo.md#detection-plus-classification-pipeline) lists every model with benchmarks.
+
+## Step 4: Scale up and explore
+
+Real surveys run many images at once. `batch_image_detection` takes a folder and a batch size, and the JSON exporters write formats downstream tools can read:
+
+```python
+results = detector.batch_image_detection("path/to/image_folder/", batch_size=16)
+```
+
+From here you can:
+
+- Run on video frames or launch the Gradio web UI, both covered in [Inference Examples](inference-examples.md).
+- Browse the full set of detectors, classifiers, and bioacoustic models in the [Model Zoo](model_zoo.md).
+- Fine-tune a model on your own data; the fine-tuning modules adapt detection and classification to your datasets.
+
+## Get help
+
+- **Email:** [zhongqimiao@microsoft.com](mailto:zhongqimiao@microsoft.com).
+- **Discord:** [the PyTorch-Wildlife community server](https://discord.gg/TeEVxzaYtm).
+
+PyTorch-Wildlife is part of the [microsoft/Biodiversity](https://github.com/microsoft/Biodiversity) umbrella; the hub links every tool in the ecosystem.

docs/index.md

@@ -22,6 +22,8 @@ tags:
 
 The goal is a shared foundation that conservation scientists can build on together: common model interfaces, reusable training and inference code, and a place to publish new architectures so the whole community benefits. Every modality-focused project in our ecosystem plugs into this framework rather than reinventing it.
 
+New here? The [Getting Started](getting-started.md) guide walks you from install to your first detection and species classification step by step.
+
 ## Why a framework, not just a model
 
 A single detection model solves one problem. Real conservation pipelines need detection, classification, batch processing, video support, and exportable results that downstream tools can read. PyTorch-Wildlife packages all of that behind one import:

docs/inference-examples.md

@@ -13,7 +13,7 @@ tags:
 
 PyTorch-Wildlife ships runnable demo scripts so you can go from a fresh install to real output without writing your own harness first. This page walks through the common ways to run inference: one image, a whole folder, a video, and an interactive web UI. Each example maps to a script under the repository's `demo/` directory.
 
-Before you start, install the framework with the [installation guide](installation.md) and skim the [API overview](api.md) so the method names below feel familiar.
+Before you start, install the framework with the [installation guide](installation.md) and skim the [API overview](api.md) so the method names below feel familiar. New to the framework? The [Getting Started](getting-started.md) guide covers the path up to this point.
 
 > [!TIP]
 > The demo scripts are written in cell style (`#%%`), so they run top to bottom as a plain `python demo/<script>.py`, or block by block in an interactive window.

docs/installation.md

@@ -11,7 +11,7 @@ tags:
 
 # Installation
 
-PyTorch-Wildlife installs as a single PyPI package, `PytorchWildlife`, on Windows, macOS, and Linux. The fastest path is a one-line `pip install`; the sections below add GPU acceleration, Docker, and platform-specific notes. Once it is installed, the [API overview](api.md) and [inference examples](inference-examples.md) show what to do next.
+PyTorch-Wildlife installs as a single PyPI package, `PytorchWildlife`, on Windows, macOS, and Linux. The fastest path is a one-line `pip install`; the sections below add GPU acceleration, Docker, and platform-specific notes. Once it is installed, the [API overview](api.md) and [inference examples](inference-examples.md) show what to do next. New to the framework? The [Getting Started](getting-started.md) guide puts this install step in context.
 
 ## Prerequisites
 

mkdocs.yml

@@ -72,6 +72,7 @@ theme:
 nav:
   - PyTorch-Wildlife:
     - Overview: index.md
+    - Getting Started: getting-started.md
     - Installation: installation.md
     - Model Zoo: model_zoo.md
     - API Overview: api.md

overrides/main.html

@@ -73,6 +73,44 @@
   </script>
   {%- endif %}
 
+  {%- if page and page.file and page.file.src_path == "getting-started.md" and page.canonical_url %}
+  {#- Getting Started page only: HowTo. Steps mirror the visible getting-started.md headings. -#}
+  <script type="application/ld+json">
+  {{- {
+    "@context": "https://schema.org",
+    "@type": "HowTo",
+    "name": "Getting Started with PyTorch-Wildlife",
+    "description": "How to get started with PyTorch-Wildlife: install the framework, run your first MegaDetector detection, add a species classifier, and scale up to batch inference.",
+    "step": [
+      {
+        "@type": "HowToStep",
+        "name": "Install the framework",
+        "text": "Install PyTorch-Wildlife from PyPI with pip, optionally in a clean conda environment. A CUDA GPU is optional, or try the hosted Hugging Face demo and Google Colab notebook without installing.",
+        "url": page.canonical_url ~ "#step-1-install-the-framework"
+      },
+      {
+        "@type": "HowToStep",
+        "name": "Run your first detection",
+        "text": "Load MegaDetectorV6 and run single_image_detection on one image. Weights download on first use.",
+        "url": page.canonical_url ~ "#step-2-run-your-first-detection"
+      },
+      {
+        "@type": "HowToStep",
+        "name": "Add species classification",
+        "text": "Pass each detection crop to a species classifier. Detection and classification share the same API, so chaining them is direct.",
+        "url": page.canonical_url ~ "#step-3-add-species-classification"
+      },
+      {
+        "@type": "HowToStep",
+        "name": "Scale up and explore",
+        "text": "Run batch_image_detection over a folder, process video or the Gradio web UI, browse the model zoo, or fine-tune on your own data.",
+        "url": page.canonical_url ~ "#step-4-scale-up-and-explore"
+      }
+    ]
+  } | tojson -}}
+  </script>
+  {%- endif %}
+
   {%- if page and page.file and page.file.src_path != "index.md" and page.canonical_url %}
   <script type="application/ld+json">
   {{- {