Fasrecon : Your Daily Fashion Recommendation

Project Plan · Project Brief · Presentation Slide · Link Video

©C242-PS216 Bangkit Capstone Team

Background

Choosing the right clothes can be a daily challenge. Factors like weather, occasion, and personal style all play a role in our decisions. Indoor or outdoor activities and sudden weather changes can make this process even more complicated. We want to make this easier by creating an android application that can provide smart clothing recommendations based on these variables. This application will integrate machine learning to evaluate user preferences, activity plans, and data from user input to provide personalized clothes recommendations. In order to get customized clothes suggestions, users may submit pictures of their clothing to the application, then the clothing classification feature will automatically detect the type and color of the clothes and the application will keep a list of the clothes the user owns. So, this can make it easier for users to manage the clothes they have. When there are clothes that are no longer used, users can delete them from the clothing collection list. After that, users can ask for clothing recommendations using a chatbot. The chatbot will provide recommendations in the form of text and also display images of clothing owned by the user according to the recommended type of clothing. This can help users who are confused about choosing the right clothes. This innovative application will not only help users look their best but also ensure they are dressed appropriately for any situation.

Team Members

ID	Learning Path	Name	University	LinkedIn
M324B4KY1763	Machine Learning	Hendri Agustono	Universitas Tanjungpura	Link
M247B4KY2802	Machine Learning	Muhammad Fadhil Syahputra	Universitas Lambung Mangkurat	Link
M324B4KY0857	Machine Learning	Bintang Budi Pangestu	Universitas Tanjungpura	Link
C529B4KY4512	Cloud Computing	Wympi Saristo	Politeknik Negeri Pontianak	Link
C529B4KY0537	Cloud Computing	Andyka Fajar Pratama	Politeknik Negeri Pontianak	Link
A529B4KY2792	Mobile Development	Muhammad Dzauqi Ikhsan	Politeknik Negeri Pontianak	Link
A324B4KY4350	Mobile Development	Tirta Kusuma Yudha	Universitas Tanjungpura	Link

User Interface

Light Theme

Dark Theme

Machine Learning Documentation

dataset link : Dataset

Mobile Development Documentation

Figma link : Figma

Application link : Application

Cloud Computing Documentation

Cloud Architecture

Endpoint

• Image Image Classification
  • https://model-classification-660807228942.asia-southeast2.run.app/predict
• Chatbot
  • https://nlp-model-660807228942.asia-southeast2.run.app/predict

Image Classification

• URL

  • /predict

• Method

  • POST

• Request Body

  • form-data
    • images : file (multiple files can be uploaded)

• Example Response

  [
      [
          "Red",
          "Tshirts"
      ],
      [
          "Blue",
          "Tshirts"
      ]
  ]
  

Chatbot

• URL

  • /predict

• Method

  • POST

• Request Body

  • raw

    • {
        "text" : "I need to go on rainy day"
      }
      

• Example Response

  {
    "outfit_recommendations": {
        "recommended_colors": [
            "Black",
            "Navy Blue"
        ],
        "recommended_outfits": [
            "Sweatshirts",
            "Jeans"
        ]
    },
    "predicted_label": "rainy",
    "recommendation_text": "Based on the rainy weather, we recommend wearing: Sweatshirts, Jeans in colors like Black, Navy Blue."
  }