{
  "id": 307048,
  "title": "🔥 Projector Visualisation Of Clothing Images + Product name",
  "url": "/competitions/h-and-m-personalized-fashion-recommendations/discussion/307048",
  "author_name": "",
  "post_date": "2022-02-12T09:12:18.063578Z",
  "votes": 26,
  "comment_count": 4,
  "views": 0,
  "content": "<p>I have created a projector visualisation of a few hundred images and text in this dataset to help with visualising this dataset. </p>\n<p>You can check out the projector visualisation <a href=\"https://cloud.relevance.ai/dataset/clothes/deploy/projector/334fe5fb667b3a64dada/dGZJazdYNEJhUXlWSV9WXzY0NUI6aWlGRlFDaExSdzY4WGJGazB5VXRoZw/JIDu7H4BMFObnPsmfsoj/us-east-1/?utm_source=kaggle\" target=\"_blank\">by following this link</a>. </p>\n<p>I found it helps a lot with EDA when you are working with a lot of images! </p>\n<p>Process I followed: </p>\n<ul>\n<li>I downloaded the data and hosted part of it online including images </li>\n<li>I vectorised all the data with OpenAI's CLIP using VectorHub </li>\n<li>I then ran PCA dimensionality reduction on the dataset and then projected it! </li>\n</ul>\n<p>We have launched! You can check out a Colab Notebook on how to build something similar in less than 5 lines of code: <a href=\"https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\" target=\"_blank\">https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja</a></p>\n<p>Also please let me know if there's any other way I can help with this :) Happy to take any feedback or feel free to leave suggestions on how the projector visualisation can be improved and I will try get back to you guys as soon as I can. </p>\n<p>Thanks! </p>",
  "messages": [
    {
      "id": "1686680",
      "postDate": "02/12/2022 09:12:18",
      "content": "<p>I have created a projector visualisation of a few hundred images and text in this dataset to help with visualising this dataset. </p>\n<p>You can check out the projector visualisation <a href=\"https://cloud.relevance.ai/dataset/clothes/deploy/projector/334fe5fb667b3a64dada/dGZJazdYNEJhUXlWSV9WXzY0NUI6aWlGRlFDaExSdzY4WGJGazB5VXRoZw/JIDu7H4BMFObnPsmfsoj/us-east-1/?utm_source=kaggle\" target=\"_blank\">by following this link</a>. </p>\n<p>I found it helps a lot with EDA when you are working with a lot of images! </p>\n<p>Process I followed: </p>\n<ul>\n<li>I downloaded the data and hosted part of it online including images </li>\n<li>I vectorised all the data with OpenAI's CLIP using VectorHub </li>\n<li>I then ran PCA dimensionality reduction on the dataset and then projected it! </li>\n</ul>\n<p>We have launched! You can check out a Colab Notebook on how to build something similar in less than 5 lines of code: <a href=\"https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\" target=\"_blank\">https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja</a></p>\n<p>Also please let me know if there's any other way I can help with this :) Happy to take any feedback or feel free to leave suggestions on how the projector visualisation can be improved and I will try get back to you guys as soon as I can. </p>\n<p>Thanks! </p>",
      "rawMarkdown": "I have created a projector visualisation of a few hundred images and text in this dataset to help with visualising this dataset. \n\nYou can check out the projector visualisation [by following this link](https://cloud.relevance.ai/dataset/clothes/deploy/projector/334fe5fb667b3a64dada/dGZJazdYNEJhUXlWSV9WXzY0NUI6aWlGRlFDaExSdzY4WGJGazB5VXRoZw/JIDu7H4BMFObnPsmfsoj/us-east-1/?utm_source=kaggle). \n\nI found it helps a lot with EDA when you are working with a lot of images! \n\nProcess I followed: \n- I downloaded the data and hosted part of it online including images \n- I vectorised all the data with OpenAI's CLIP using VectorHub \n- I then ran PCA dimensionality reduction on the dataset and then projected it! \n\nWe have launched! You can check out a Colab Notebook on how to build something similar in less than 5 lines of code: https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\n\nAlso please let me know if there's any other way I can help with this :) Happy to take any feedback or feel free to leave suggestions on how the projector visualisation can be improved and I will try get back to you guys as soon as I can. \n\nThanks!",
      "votes": null
    },
    {
      "id": "1690450",
      "postDate": "02/15/2022 00:35:16",
      "content": "<p>Added a Colab notebook to show others how to build their own in less than 5 lines of code! </p>\n<p><a href=\"https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\" target=\"_blank\">https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja</a></p>",
      "rawMarkdown": "Added a Colab notebook to show others how to build their own in less than 5 lines of code! \n\nhttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja",
      "votes": null
    },
    {
      "id": "1694723",
      "postDate": "02/17/2022 17:02:22",
      "content": "<p>Really cool stuff. You can also paste a snapshot of the projection to generate more interest. First time hearing about Relevance AI, seems interesting.</p>",
      "rawMarkdown": "Really cool stuff. You can also paste a snapshot of the projection to generate more interest. First time hearing about Relevance AI, seems interesting.",
      "votes": null
    },
    {
      "id": "1695116",
      "postDate": "02/18/2022 00:55:08",
      "content": "<p>Very cool, thanks</p>",
      "rawMarkdown": "Very cool, thanks",
      "votes": null
    },
    {
      "id": "1699227",
      "postDate": "02/21/2022 03:17:34",
      "content": "<p>I tried but unfortunately Kaggle's image-pasting is bugged out! Waiting for a fix :) </p>",
      "rawMarkdown": "I tried but unfortunately Kaggle's image-pasting is bugged out! Waiting for a fix :)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1690450,
      "author_name": "jacky2wong",
      "author_url": "",
      "post_date": "02/15/2022 00:35:16",
      "content": "<p>Added a Colab notebook to show others how to build their own in less than 5 lines of code! </p>\n<p><a href=\"https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\" target=\"_blank\">https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja</a></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1694723,
      "author_name": "aussie84",
      "author_url": "",
      "post_date": "02/17/2022 17:02:22",
      "content": "<p>Really cool stuff. You can also paste a snapshot of the projection to generate more interest. First time hearing about Relevance AI, seems interesting.</p>",
      "votes": null,
      "replies": [
        {
          "id": 1699227,
          "author_name": "jacky2wong",
          "author_url": "",
          "post_date": "02/21/2022 03:17:34",
          "content": "<p>I tried but unfortunately Kaggle's image-pasting is bugged out! Waiting for a fix :) </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1695116,
      "author_name": "cdeotte",
      "author_url": "",
      "post_date": "02/18/2022 00:55:08",
      "content": "<p>Very cool, thanks</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1686680": "I have created a projector visualisation of a few hundred images and text in this dataset to help with visualising this dataset. \n\nYou can check out the projector visualisation [by following this link](https://cloud.relevance.ai/dataset/clothes/deploy/projector/334fe5fb667b3a64dada/dGZJazdYNEJhUXlWSV9WXzY0NUI6aWlGRlFDaExSdzY4WGJGazB5VXRoZw/JIDu7H4BMFObnPsmfsoj/us-east-1/?utm_source=kaggle). \n\nI found it helps a lot with EDA when you are working with a lot of images! \n\nProcess I followed: \n- I downloaded the data and hosted part of it online including images \n- I vectorised all the data with OpenAI's CLIP using VectorHub \n- I then ran PCA dimensionality reduction on the dataset and then projected it! \n\nWe have launched! You can check out a Colab Notebook on how to build something similar in less than 5 lines of code: https://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja\n\nAlso please let me know if there's any other way I can help with this :) Happy to take any feedback or feel free to leave suggestions on how the projector visualisation can be improved and I will try get back to you guys as soon as I can. \n\nThanks!",
    "1690450": "Added a Colab notebook to show others how to build their own in less than 5 lines of code! \n\nhttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharinghttps://colab.research.google.com/drive/12GiCo9r0JpFqcg5Tmae28Wz0VPMpHul_?usp=sharing?utm_source=kaggle_wongja",
    "1694723": "Really cool stuff. You can also paste a snapshot of the projection to generate more interest. First time hearing about Relevance AI, seems interesting.",
    "1695116": "Very cool, thanks",
    "1699227": "I tried but unfortunately Kaggle's image-pasting is bugged out! Waiting for a fix :)"
  },
  "source": "meta"
}