change YT video embedding

Author: elenaivadreyer

tutorial.ipynb

@@ -127,27 +127,54 @@
     "\n",
     "### Key Concepts\n",
     "\n",
-    "- **Few-Shot Learning:** Few-shot learning aims to train models that can recognize new classes from only a handful of labeled examples by leveraging prior knowledge learned from many related tasks.\n",
-    "\n",
-    "<div style=\"text-align: center;\">\n",
-    "  <iframe\n",
-    "    width=\"560\" height=\"315\"\n",
-    "    src=\"https://www.youtube.com/embed/hE7eGew4eeg\"\n",
-    "    title=\"YouTube video player\"\n",
-    "    frameborder=\"0\"\n",
-    "    allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture\"\n",
-    "    allowfullscreen>\n",
-    "  </iframe>\n",
-    "</div>\n",
-    "\n",
+    "- **Few-Shot Learning:** Few-shot learning aims to train models that can recognize new classes from only a handful of labeled examples by leveraging prior knowledge learned from many related tasks."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "\n",
+       "        <iframe\n",
+       "            width=\"560\"\n",
+       "            height=\"315\"\n",
+       "            src=\"https://www.youtube.com/embed/hE7eGew4eeg\"\n",
+       "            frameborder=\"0\"\n",
+       "            allowfullscreen\n",
+       "            \n",
+       "        ></iframe>\n",
+       "        "
+      ],
+      "text/plain": [
+       "<IPython.lib.display.IFrame at 0x13beb6710>"
+      ]
+     },
+     "execution_count": 13,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "from IPython.display import IFrame\n",
     "\n",
+    "IFrame(\"https://www.youtube.com/embed/hE7eGew4eeg\", width=560, height=315)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
     "- **Semantic Segmentation:** In the context of semantic segmentation, this means learning to assign a class label to every pixel in an image, even when only a few annotated images are available for each class.\n",
     "\n",
     "- **Prototypical Networks:** Prototypical Networks tackle this by mapping pixels (or patches) into an embedding space where each class is represented by a prototype, typically the mean embedding of its support examples. At inference time, each pixel in a query image is classified by measuring its distance to these class prototypes, enabling accurate pixel-wise segmentation in new scenes with very limited labeled data.\n",
     "\n",
     "![](presentation/figures/illustration_prototypical_network.png)\n",
     "\n",
-    "*Figure 1: Overview of the Self-Regularized Prototypical Network Architecture. Adapted from [Ding et al. (2022)](https://arxiv.org/pdf/2210.16829).*\n"
+    "*Figure 1: Overview of the Self-Regularized Prototypical Network Architecture. Adapted from [Ding et al. (2022)](https://arxiv.org/pdf/2210.16829).*"
    ]
   },
   {
@@ -173,7 +200,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -196,6 +223,8 @@
     "import sys\n",
     "import os\n",
     "\n",
+    "os.getcwd()\n",
+    "\n",
     "# Add parent directory to sys.path to allow imports from there\n",
     "sys.path.append(os.path.abspath(\"..\"))\n",
     "\n",
@@ -237,7 +266,7 @@
    "source": [
     "## Data Download\n",
     "\n",
-    "The dataset is stored on the Hugging Face Hub and each tile comes with:\n",
+    "The dataset is stored on the [Hugging Face Hub](https://huggingface.co/datasets/raphaelattias/overfitteam-geneva-satellite-images) and each tile comes with:\n",
     "\n",
     "- a **RGB satellite image**, and\n",
     "- a **binary mask** indicating pixels that belong to rooftops suitable for PV installations.\n",