Merge pull request fastai#139 from appleby/lesson7-download-annotations

Add get_annotations and skip 'other' when loading anno_classes
chriii · Aug 2, 2017 · 8e64556 · 8e64556
2 parents 4dc23e5 + 367e2b7
commit 8e64556
Showing 1 changed file with 54 additions and 0 deletions.
diff --git a/deeplearning1/nbs/lesson7.ipynb b/deeplearning1/nbs/lesson7.ipynb
@@ -1079,6 +1079,59 @@
     "anno_classes = ['alb', 'bet', 'dol', 'lag', 'other', 'shark', 'yft']"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {
+    "collapsed": true
+   },
+   "outputs": [],
+   "source": [
+    "def get_annotations():\n",
+    "    annot_urls = {\n",
+    "        '5458/bet_labels.json': 'bd20591439b650f44b36b72a98d3ce27',\n",
+    "        '5459/shark_labels.json': '94b1b3110ca58ff4788fb659eda7da90',\n",
+    "        '5460/dol_labels.json': '91a25d29a29b7e8b8d7a8770355993de',\n",
+    "        '5461/yft_labels.json': '9ef63caad8f076457d48a21986d81ddc',\n",
+    "        '5462/alb_labels.json': '731c74d347748b5272042f0661dad37c',\n",
+    "        '5463/lag_labels.json': '92d75d9218c3333ac31d74125f2b380a'\n",
+    "    }\n",
+    "    cache_subdir = os.path.abspath(os.path.join(path, 'annos'))\n",
+    "    url_prefix = 'https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/'\n",
+    "    \n",
+    "    if not os.path.exists(cache_subdir):\n",
+    "        os.makedirs(cache_subdir)\n",
+    "    \n",
+    "    for url_suffix, md5_hash in annot_urls.iteritems():\n",
+    "        fname = url_suffix.rsplit('/', 1)[-1]\n",
+    "        get_file(fname, url_prefix + url_suffix, cache_subdir=cache_subdir, md5_hash=md5_hash)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Downloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5463/lag_labels.json\n",
+      "16384/30731 [==============>...............] - ETA: 0sDownloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5459/shark_labels.json\n",
+      "49152/68097 [====================>.........] - ETA: 0sDownloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5461/yft_labels.json\n",
+      "212992/284511 [=====================>........] - ETA: 0sDownloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5458/bet_labels.json\n",
+      "49152/82471 [================>.............] - ETA: 0sDownloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5462/alb_labels.json\n",
+      "606208/775061 [======================>.......] - ETA: 0sDownloading data from https://kaggle2.blob.core.windows.net/forum-message-attachments/147157/5460/dol_labels.json\n",
+      "16384/41584 [==========>...................] - ETA: 0s"
+     ]
+    }
+   ],
+   "source": [
+    "get_annotations()"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 48,
@@ -1089,6 +1142,7 @@
    "source": [
     "bb_json = {}\n",
     "for c in anno_classes:\n",
+    "    if c == 'other': continue # no annotation file for \"other\" class\n",
     "    j = json.load(open('{}annos/{}_labels.json'.format(path, c), 'r'))\n",
     "    for l in j:\n",
     "        if 'annotations' in l.keys() and len(l['annotations'])>0:\n",