video/text_video_retrieval/1_text_video_retrieval_engine.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "868d7f57",
   "metadata": {},
   "source": [
    "# How to Build a Text-Video Retrieval Engine\n",
    "\n",
    "This notebook illustrates how to build a text-video retrieval engine from scratch using [Milvus](https://milvus.io/) and [Towhee](https://towhee.io/).\n",
    "\n",
    "\n",
    "**What is Text-Video Retrieval?**\n",
    "\n",
    "In simple words, text-video retrieval is: given a text query and a pool of candidate videos, select the video which corresponds to the text query.\n",
    "\n",
    "\n",
    "**What are Milvus & Towhee?**\n",
    "\n",
    "- Milvus is the most advanced open-source vector database built for AI applications and supports nearest neighbor embedding search across tens of millions of entries.\n",
    "- Towhee is a framework that provides ETL for unstructured data using SoTA machine learning models.\n",
    "\n",
    "We'll go through video retrieval procedures and evaluate the performance. Moreover, we managed to make the core functionality as simple as few lines of code, with which you can start hacking your own video retrieval engine.\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "157044a8",
   "metadata": {},
   "source": [
    "## Preparation\n",
    "\n",
    "### Install packages\n",
    "\n",
    "Make sure you have installed required python packages:\n",
    "\n",
    "| package |\n",
    "| -- |\n",
    "| pymilvus |\n",
    "| towhee |\n",
    "| towhee.models |\n",
    "| pillow |\n",
    "| ipython |\n",
    "| gradio |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "4bc9560b",
   "metadata": {},
   "outputs": [],
   "source": [
    "! python -m pip install -q pymilvus towhee towhee.models pillow ipython gradio"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5891a4e6",
   "metadata": {},
   "source": [
    "### Prepare the data\n",
    "\n",
    "First, we need to prepare the dataset and Milvus environment.   \n",
    "\n",
    "[MSR-VTT (Microsoft Research Video to Text)](https://www.microsoft.com/en-us/research/publication/msr-vtt-a-large-video-description-dataset-for-bridging-video-and-language/) is a dataset for the open domain video captioning, which consists of 10,000 video clips.  \n",
    "\n",
    "Download the MSR-VTT-1kA test set from [google drive](https://drive.google.com/file/d/1cuFpHiK3jV9cZDKcuGienxTg1YQeDs-w/view?usp=sharing) and unzip it, which contains just 1k videos.  \n",
    "And the video captions text sentence information is in ./MSRVTT_JSFUSION_test.csv.\n",
    "\n",
    "The data is organized as follows:\n",
    "- **test_1k_compress:** 1k compressed test videos in MSR-VTT-1kA.\n",
    "- **MSRVTT_JSFUSION_test.csv:** a csv file containing an ***key,vid_key,video_id,sentence***, for each video and caption text.\n",
    "\n",
    "Let's take a quick look"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "478b8769",
   "metadata": {},
   "outputs": [],
   "source": [
    "! curl -L https://github.com/towhee-io/examples/releases/download/data/text_video_search.zip -O\n",
    "! unzip -q -o text_video_search.zip"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "23d0e3a6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "length of all test set is 1000\n",
      "random sample 1000 examples\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>video_id</th>\n",
       "      <th>video_path</th>\n",
       "      <th>sentence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>video7579</td>\n",
       "      <td>./test_1k_compress/video7579.mp4</td>\n",
       "      <td>a girl wearing red top and black trouser is pu...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>video7725</td>\n",
       "      <td>./test_1k_compress/video7725.mp4</td>\n",
       "      <td>young people sit around the edges of a room cl...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>video9258</td>\n",
       "      <td>./test_1k_compress/video9258.mp4</td>\n",
       "      <td>a person is using a phone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>video7365</td>\n",
       "      <td>./test_1k_compress/video7365.mp4</td>\n",
       "      <td>cartoon people are eating at a restaurant</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>video8068</td>\n",
       "      <td>./test_1k_compress/video8068.mp4</td>\n",
       "      <td>a woman on a couch talks to a a man</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    video_id                        video_path  \\\n",
       "0  video7579  ./test_1k_compress/video7579.mp4   \n",
       "1  video7725  ./test_1k_compress/video7725.mp4   \n",
       "2  video9258  ./test_1k_compress/video9258.mp4   \n",
       "3  video7365  ./test_1k_compress/video7365.mp4   \n",
       "4  video8068  ./test_1k_compress/video8068.mp4   \n",
       "\n",
       "                                            sentence  \n",
       "0  a girl wearing red top and black trouser is pu...  \n",
       "1  young people sit around the edges of a room cl...  \n",
       "2                          a person is using a phone  \n",
       "3          cartoon people are eating at a restaurant  \n",
       "4                a woman on a couch talks to a a man  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "raw_video_path = './test_1k_compress' # 1k test video path.\n",
    "test_csv_path = './MSRVTT_JSFUSION_test.csv' # 1k video caption csv.\n",
    "\n",
    "test_sample_csv_path = './MSRVTT_JSFUSION_test_sample.csv'\n",
    "\n",
    "sample_num = 1000 # you can change this sample_num to be smaller, so that this notebook will be faster.\n",
    "test_df = pd.read_csv(test_csv_path)\n",
    "print('length of all test set is {}'.format(len(test_df)))\n",
    "sample_df = test_df.sample(sample_num, random_state=42)\n",
    "\n",
    "sample_df['video_path'] = sample_df.apply(lambda x:os.path.join(raw_video_path, x['video_id']) + '.mp4', axis=1)\n",
    "\n",
    "sample_df.to_csv(test_sample_csv_path)\n",
    "print('random sample {} examples'.format(sample_num))\n",
    "\n",
    "df = pd.read_csv(test_sample_csv_path)\n",
    "\n",
    "df[['video_id', 'video_path', 'sentence']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ae4911f0",
   "metadata": {},
   "source": [
    "Define some helper function to convert video to gif so that we can have a look at these video-text pairs.   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "5ad52c9c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from IPython import display\n",
    "from pathlib import Path\n",
    "import towhee\n",
    "from PIL import Image\n",
    "\n",
    "def display_gif(video_path_list, text_list):\n",
    "    html = ''\n",
    "    for video_path, text in zip(video_path_list, text_list):\n",
    "        html_line = '<img src=\"{}\"> {} <br/>'.format(video_path, text)\n",
    "        html += html_line\n",
    "    return display.HTML(html)\n",
    "\n",
    "    \n",
    "def convert_video2gif(video_path, output_gif_path, num_samples=16):\n",
    "    frames = (\n",
    "        towhee.glob(video_path)\n",
    "              .video_decode.ffmpeg(sample_type='uniform_temporal_subsample', args={'num_samples': num_samples})\n",
    "              .to_list()[0]\n",
    "    )\n",
    "    imgs = [Image.fromarray(frame) for frame in frames]\n",
    "    imgs[0].save(fp=output_gif_path, format='GIF', append_images=imgs[1:], save_all=True, loop=0)\n",
    "\n",
    "\n",
    "def display_gifs_from_video(video_path_list, text_list, tmpdirname = './tmp_gifs'):\n",
    "    Path(tmpdirname).mkdir(exist_ok=True)\n",
    "    gif_path_list = []\n",
    "    for video_path in video_path_list:\n",
    "        video_name = str(Path(video_path).name).split('.')[0]\n",
    "        gif_path = Path(tmpdirname) / (video_name + '.gif')\n",
    "        convert_video2gif(video_path, gif_path)\n",
    "        gif_path_list.append(gif_path)\n",
    "    return display_gif(gif_path_list, text_list)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7bcdea96",
   "metadata": {},
   "source": [
    "Take a look at the ground-truth video-text pairs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5dbb2210",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"tmp_gifs/video7579.gif\"> a girl wearing red top and black trouser is putting a sweater on a dog <br/><img src=\"tmp_gifs/video7725.gif\"> young people sit around the edges of a room clapping and raising their arms while others dance in the center during a party <br/><img src=\"tmp_gifs/video9258.gif\"> a person is using a phone <br/><img src=\"tmp_gifs/video7365.gif\"> cartoon people are eating at a restaurant <br/><img src=\"tmp_gifs/video8068.gif\"> a woman on a couch talks to a a man <br/>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sample_show_df = sample_df.sample(5, random_state=42)\n",
    "sample_show_df = sample_df[:5]\n",
    "video_path_list = sample_show_df['video_path'].to_list()\n",
    "text_list = sample_show_df['sentence'].to_list()\n",
    "tmpdirname = './tmp_gifs'\n",
    "display_gifs_from_video(video_path_list, text_list, tmpdirname=tmpdirname)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c57038b4",
   "metadata": {},
   "source": [
    "### Create a Milvus Collection\n",
    "\n",
    "Before getting started, please make sure you have [installed milvus](https://milvus.io/docs/v2.0.x/install_standalone-docker.md). Let's first create a `video retrieval` collection that uses the [L2 distance metric](https://milvus.io/docs/v2.0.x/metric.md#Euclidean-distance-L2) and an [IVF_FLAT index](https://milvus.io/docs/v2.0.x/index.md#IVF_FLAT)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "84807c03",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pymilvus import connections, FieldSchema, CollectionSchema, DataType, Collection, utility\n",
    "\n",
    "connections.connect(host='127.0.0.1', port='19530')\n",
    "\n",
    "def create_milvus_collection(collection_name, dim):\n",
    "    if utility.has_collection(collection_name):\n",
    "        utility.drop_collection(collection_name)\n",
    "    \n",
    "    fields = [\n",
    "    FieldSchema(name='id', dtype=DataType.INT64, descrition='ids', is_primary=True, auto_id=False),\n",
    "    FieldSchema(name='embedding', dtype=DataType.FLOAT_VECTOR, descrition='embedding vectors', dim=dim)\n",
    "    ]\n",
    "    schema = CollectionSchema(fields=fields, description='video retrieval')\n",
    "    collection = Collection(name=collection_name, schema=schema)\n",
    "\n",
    "    # create IVF_FLAT index for collection.\n",
    "    index_params = {\n",
    "        'metric_type':'L2', #IP\n",
    "        'index_type':\"IVF_FLAT\",\n",
    "        'params':{\"nlist\":2048}\n",
    "    }\n",
    "    collection.create_index(field_name=\"embedding\", index_params=index_params)\n",
    "    return collection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c2798b8d",
   "metadata": {},
   "outputs": [],
   "source": [
    "collection = create_milvus_collection('text_video_retrieval', 512)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7c72d0e8",
   "metadata": {},
   "source": [
    "## Text-Video retrieval\n",
    "\n",
    "In this section, we'll show how to build our text-video retrieval engine using Milvus. The basic idea behind text-video retrieval is the extract embeddings from videos using a Transformer network and store them in Milvus, then using another Transformer network to get text embeddings and compare with those stored in Milvus.\n",
    "\n",
    "We use [Towhee](https://towhee.io/), a machine learning framework that allows for creating data processing pipelines. [Towhee](https://towhee.io/) also provides predefined operators which implement insert and query operation in Milvus.\n"
   ]
  },
  {
   "attachments": {
    "image-4.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "f5326800",
   "metadata": {},
   "source": [
    "### Load Video Embeddings into Milvus\n",
    "\n",
    "We first extract embeddings from images with `CLIP4Clip` model and insert the embeddings into Milvus for indexing. Towhee provides a [method-chaining style API](https://towhee.readthedocs.io/en/main/index.html) so that users can assemble a data processing pipeline with operators.   \n",
    "\n",
    "[CLIP4Clip](https://arxiv.org/abs/2104.08860) is a video-text retrieval model based on [CLIP (ViT-B)](https://github.com/openai/CLIP). The [towhee clip4clip operator](https://towhee.io/video-text-embedding/clip4clip) with pretrained weights can easily extract video embedding and text embedding by a few codes.\n",
    "\n",
    "![image-4.png](attachment:image-4.png)\n",
    "\n",
    "Before you start running the clip4clip operator, you should make sure you have make [git](https://git-scm.com/) and [git-lfs](https://git-lfs.github.com/) installed.    \n",
    "\n",
    "For git(If you have installed, just skip it):\n",
    "```\n",
    "sudo apt-get install git\n",
    "```\n",
    "For git-lfs(You must install it for downloading checkpoint weights on backend):\n",
    "```\n",
    "sudo apt-get install git-lfs\n",
    "git lfs install\n",
    "```\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "a703bd54",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 14min 30s, sys: 1min 55s, total: 1h 16min 25s\n",
      "Wall time: 1min 19s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "import os\n",
    "import towhee\n",
    "\n",
    "device = 'cuda:2'\n",
    "# device = 'cpu'\n",
    "\n",
    "# For the first time you run this line, \n",
    "# it will take some time \n",
    "# because towhee will download operator with weights on backend.\n",
    "dc = (\n",
    "    towhee.read_csv(test_sample_csv_path).unstream()\n",
    "      .runas_op['video_id', 'id'](func=lambda x: int(x[-4:]))\n",
    "      .video_decode.ffmpeg['video_path', 'frames'](sample_type='uniform_temporal_subsample', args={'num_samples': 12}) \\\n",
    "      .runas_op['frames', 'frames'](func=lambda x: [y for y in x]) \\\n",
    "      .video_text_embedding.clip4clip['frames', 'vec'](model_name='clip_vit_b32', modality='video', device=device) \\\n",
    "      .to_milvus['id', 'vec'](collection=collection, batch=30)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "7dba693e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total number of inserted data is 1000.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">video_id</th> <th style=\"text-align: center; font-size: 130%; border: none;\">id</th> <th style=\"text-align: center; font-size: 130%; border: none;\">vec</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7579</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">7579</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[-0.046955414, -0.012318517, 0.05594952, ...] shape=(512,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7725</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">7725</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[0.057752915, 0.026032623, -0.017264651, ...] shape=(512,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video9258</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">9258</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[-0.01112327, 0.054949787, -0.033538133, ...] shape=(512,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7365</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">7365</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[-0.0137340445, -0.0064378446, -0.02168003, ...] shape=(512,)</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video8068</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">8068</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[0.023609513, -0.033269204, 0.0024328423, ...] shape=(512,)</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('Total number of inserted data is {}.'.format(collection.num_entities))\n",
    "dc.select['video_id', 'id', 'vec']().show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6bb5e044",
   "metadata": {},
   "source": [
    "Here is detailed explanation for each line of the code:\n",
    "\n",
    "- `towhee.read_csv(test_sample_csv_path)`: read tabular data from csv file;\n",
    "\n",
    "- `.runas_op['video_id', 'id'](func=lambda x: int(x[-4:]))`: for each row from the data, convert the data type of the column `id` from last 4 number of `video_id`;\n",
    "\n",
    "- `.video_decode.ffmpeg` and `runas_op`: subsample the video uniformly, and then get a list of images in the video, which are the input of the clip4clip model;\n",
    "\n",
    "- `.video_text_embedding.clip4clip['frames', 'vec'](model_name='clip_vit_b32', modality='video')`: extract embedding feature from the images subsampled from video, and then mean pool them in the temporal dimension, which repre.\n",
    "\n",
    "- `.to_milvus['id', 'vec'](collection=collection, batch=30)`: insert video embedding features in to Milvus;\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3da5850f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 47min 32s, sys: 2min 37s, total: 1h 50min 10s\n",
      "Wall time: 2min 2s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "dc = (\n",
    "    towhee.read_csv(test_sample_csv_path).unstream()\n",
    "      .video_text_embedding.clip4clip['sentence','text_vec'](model_name='clip_vit_b32', modality='text', device=device)\n",
    "      .milvus_search['text_vec', 'top10_raw_res'](collection=collection, limit=10)\n",
    "      .runas_op['video_id', 'ground_truth'](func=lambda x : [int(x[-4:])])\n",
    "      .runas_op['top10_raw_res', 'top1'](func=lambda res: [x.id for i, x in enumerate(res) if i < 1])\n",
    "      .runas_op['top10_raw_res', 'top5'](func=lambda res: [x.id for i, x in enumerate(res) if i < 5])\n",
    "      .runas_op['top10_raw_res', 'top10'](func=lambda res: [x.id for i, x in enumerate(res) if i < 10])\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bf7722c7",
   "metadata": {},
   "source": [
    "## Evaluation\n",
    "\n",
    "We have finished the core functionality of the text-video retrieval engine. However, we don't know whether it achieves a reasonable performance. We need to evaluate the retrieval engine against the ground truth so that we know if there is any room to improve it.\n",
    "\n",
    "In this section, we'll evaluate the strength of our text-video retrieval using recall@topk:   \n",
    "`Recall@topk` is the proportion of relevant items found in the top-k recommendations. Suppose that we computed recall at 10 and found it is 40% in our top-10 recommendation system. This means that 40% of the total number of the relevant items appear in the top-k results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "8a75c143",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border-collapse: collapse;\"><tr><th style=\"text-align: center; font-size: 130%; border: none;\">video_id</th> <th style=\"text-align: center; font-size: 130%; border: none;\">sentence</th> <th style=\"text-align: center; font-size: 130%; border: none;\">ground_truth</th> <th style=\"text-align: center; font-size: 130%; border: none;\">top10_raw_res</th> <th style=\"text-align: center; font-size: 130%; border: none;\">top1</th> <th style=\"text-align: center; font-size: 130%; border: none;\">top5</th> <th style=\"text-align: center; font-size: 130%; border: none;\">top10</th></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7579</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">a girl wearing red top and black...</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7579] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[{\"id\": 7579, \"score\": 1.4152562618255615},{\"id\": 9969, \"score\": 1.479762315750122},{\"id\": 8837, \"score\": 1.4900273084640503},{\"id\": 9347, \"score\": 1.4925360679626465},...] len=10</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7579] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7579,9969,8837,9347,...] len=5</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7579,9969,8837,9347,...] len=10</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7725</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">young people sit around the edge...</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7725] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[{\"id\": 7725, \"score\": 1.3606964349746704},{\"id\": 8014, \"score\": 1.490865707397461},{\"id\": 8339, \"score\": 1.4913759231567383},{\"id\": 8442, \"score\": 1.503324270248413},...] len=10</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7725] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7725,8014,8339,8442,...] len=5</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7725,8014,8339,8442,...] len=10</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video9258</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">a person is using a phone</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[9258] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[{\"id\": 9258, \"score\": 1.401209831237793},{\"id\": 9257, \"score\": 1.4216816425323486},{\"id\": 9697, \"score\": 1.4404691457748413},{\"id\": 7910, \"score\": 1.4957969188690186},...] len=10</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[9258] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[9258,9257,9697,7910,...] len=5</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[9258,9257,9697,7910,...] len=10</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video7365</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">cartoon people are eating at a r...</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7365] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[{\"id\": 7365, \"score\": 1.4048435688018799},{\"id\": 8781, \"score\": 1.4607740640640259},{\"id\": 9537, \"score\": 1.4721850156784058},{\"id\": 7831, \"score\": 1.5039441585540771},...] len=10</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7365] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7365,8781,9537,7831,...] len=5</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7365,8781,9537,7831,...] len=10</td></tr> <tr><td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">video8068</td> <td style=\"text-align: center; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">a woman on a couch talks to a a ...</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[8068] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[{\"id\": 7162, \"score\": 1.4739655256271362},{\"id\": 8304, \"score\": 1.4785505533218384},{\"id\": 8068, \"score\": 1.4936964511871338},{\"id\": 7724, \"score\": 1.495842456817627},...] len=10</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7162] len=1</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7162,8304,8068,7724,...] len=5</td> <td style=\"text-align: left; vertical-align: center; border-right: solid 1px #D3D3D3; border-left: solid 1px #D3D3D3; \">[7162,8304,8068,7724,...] len=10</td></tr></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "dc.select['video_id', 'sentence', 'ground_truth', 'top10_raw_res', 'top1', 'top5', 'top10']().show()\n",
    "# dc.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "f89a4b42",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>mean_hit_ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>recall_at_1</th>\n",
       "      <td>0.421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall_at_5</th>\n",
       "      <td>0.712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall_at_10</th>\n",
       "      <td>0.813</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              mean_hit_ratio\n",
       "recall_at_1            0.421\n",
       "recall_at_5            0.712\n",
       "recall_at_10           0.813"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "benchmark = (\n",
    "    dc.with_metrics(['mean_hit_ratio',]) \\\n",
    "        .evaluate['ground_truth', 'top1'](name='recall_at_1') \\\n",
    "        .evaluate['ground_truth', 'top5'](name='recall_at_5') \\\n",
    "        .evaluate['ground_truth', 'top10'](name='recall_at_10') \\\n",
    "        .report()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b886b88",
   "metadata": {},
   "source": [
    "This result is almost identical to the recall metrics represented in the paper. You can find more detail about metrics in [paperwithcode](https://paperswithcode.com/paper/clip4clip-an-empirical-study-of-clip-for-end/review/?hl=30331)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a1e2a48",
   "metadata": {},
   "source": [
    "## Release a Showcase\n",
    "\n",
    "We've learnt how to build a reverse video search engine. Now it's time to add some interface and release a showcase. Towhee provides `towhee.api()` to wrap the data processing pipeline as a function with `.as_function()`. So we can build a quick demo with this `milvus_search_function` with [Gradio](https://gradio.app/)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "69b74139",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running on local URL:  http://127.0.0.1:7861/\n",
      "Running on public URL: https://29428.gradio.app\n",
      "\n",
      "This share link expires in 72 hours. For free permanent hosting, check out Spaces (https://huggingface.co/spaces)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div><iframe src=\"https://29428.gradio.app\" width=\"900\" height=\"500\" allow=\"autoplay; camera; microphone;\" frameborder=\"0\" allowfullscreen></iframe></div>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(<gradio.routes.App at 0x7fd3c7033550>,\n",
       " 'http://127.0.0.1:7861/',\n",
       " 'https://29428.gradio.app')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import gradio\n",
    "\n",
    "show_num = 3\n",
    "with towhee.api() as api:\n",
    "    milvus_search_function = (\n",
    "         api.video_text_embedding.clip4clip(model_name='clip_vit_b32', modality='text', device=device)\n",
    "            .milvus_search(collection=collection, limit=show_num)\n",
    "            .runas_op(func=lambda res: [os.path.join(raw_video_path, 'video' + str(x.id) + '.mp4') for x in res])\n",
    "            .as_function()\n",
    "    )\n",
    "\n",
    "interface = gradio.Interface(milvus_search_function, \n",
    "                             inputs=[gradio.Textbox()],\n",
    "                             outputs=[gradio.Video(format='mp4') for _ in range(show_num)]\n",
    "                            )\n",
    "\n",
    "interface.launch(inline=True, share=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dcb7f906",
   "metadata": {},
   "outputs": [],
   "source": [
    "# import shutil\n",
    "# shutil.rmtree(tmpdirname)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "070f1ab7",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}