Stylumia NXT Hackathon 2024

87 Registered Allowed team size: 1 - 4
87 Registered Allowed team size: 1 - 4
idea phase
Online
starts on:
Nov 13, 2024, 12:30 PM UTC (UTC)
ends on:
Dec 15, 2024, 06:29 PM UTC (UTC)
Prototype phase
Online
starts on:
Dec 20, 2024, 12:30 PM UTC (UTC)
ends on:
Jan 05, 2025, 06:29 PM UTC (UTC)

Overview

The Stylumia NXT Hackathon invites designers, data scientists, and innovators to reimagine the next generation of intelligence for fashion retail. Retailers face crucial decisions daily, such as determining which products to include in assortments based on trends, setting optimal pricing, and planning promotions. At Stylumia, we aim to overcome industry challenges like subjective forecasting and inefficient demand planning, which lead to unsustainable practices. This hackathon provides a platform to explore new possibilities at the intersection of AI, data science, and user experience, with the goal of enabling smarter, faster, and more sustainable decisions in fashion.

Participants will tackle key challenges in trend discovery and product intelligence, focusing on creating solutions that are not only functional but also future-proof. This collaborative event encourages participants to push boundaries and develop practical tools that align with Stylumia’s mission to augment human decision-making with technology, reducing waste and improving business outcomes.

Objective

Stylumia NXT Hackathon seeks to inspire solutions that leverage AI and build next-generation user experiences to redefine the future of fashion retail, focusing on two core challenges where innovation can drive meaningful change:

  • Reimagining Trend Discovery for Designers: Designers often struggle with tools that are complex, generic, and static, limiting their ability to quickly identify and act on emerging trends. Participants will develop engaging, intuitive, and personalized interfaces to improve trend tracking and discovery. These solutions should empower designers with real-time insights, making the process of trend exploration dynamic, easy, and enjoyable.
  • Developing a Universal Fashion Ontology & Feature Extraction System: Fashion products span a wide range of categories, from apparel and footwear to accessories, each with distinct attributes. Traditional systems often fall short in capturing and analyzing these intricate product details. Participants will design AI solutions capable of extracting, organizing, and understanding product attributes with greater precision, helping retailers optimize assortments and make better decisions based on consumer demand.

Eligibility Criteria

  • Open to All

AI Problem Statement - Universal Fashion Ontology & Feature Extraction System

  • Who is it for?
    • This challenge is for AI enthusiasts, data scientists, and developers eager to create innovative solutions that enhance product discovery, trend forecasting, and personalization in fashion retail. It’s ideal for those skilled in computer vision, NLP, and scalable tech design.

UX Design Problem Statement – Designing the Future of Fashion Trend Discovery

  • Who is it for?
    • This challenge is for UX designers, product managers, and innovators passionate about transforming trend discovery for fashion professionals. It's ideal for those who want to create engaging, user-friendly B2B solutions for designers, planners, and merchandisers.

Themes

AI Problem Statement - Universal Fashion Ontology & Feature Extraction System

Overview:

Fashion spans a wide range of categories, such as apparel, footwear, and accessories, each with intricate features like materials, patterns, and designs. Traditional systems struggle to capture these complexities effectively. This theme challenges participants to create an AI-powered solution that extracts and organizes product attributes accurately, driving trend forecasting, personalized recommendations, and better retail decisions.

Challege:

Participants will develop a universal fashion ontology and feature extraction system that integrates both visual and textual data. The goal is to create a comprehensive ontology capable of accurately representing the wide spectrum of fashion features across categories. This ontology must be flexible enough to evolve with new trends while remaining structured to maintain consistency. The solution should capture nuanced product features within their specific contexts while being scalable to handle large datasets and adaptable to retail needs.

Key Consideration:

  • Multi-modal Integration: Combine visual (images) and textual (descriptions, reviews) data to extract product features with precision.
  • Ontological Framework: Develop an adaptable and structured ontology that evolves with new fashion trends and products.
  • Contextual Understanding: Ensure the system interprets features appropriately within the context of different product categories (e.g., "cold shoulder" in a dress vs. a sweater).
  • Scalability and Precision: Balance the ability to scale across large datasets with the need for precise, nuanced feature extraction.

Submission Package Requirements:

  1. Documentation Format
    • Technical documentation in PDF format
    • Architecture diagrams in standard formats (SVG/PNG)
    • Interactive demonstrations (if applicable)
  2. Code Repository
    • Private GitHub/GitLab repository with access provided
    • README.md with clear setup instructions
    • Requirements.txt or equivalent dependency file
    • Code comments and documentation
  3. Presentation Materials
    • Executive summary (2 pages max)
    • Technical presentation slides (20 slides max)
    • Demo video (5-10 minutes)
    • Interactive notebook demonstrations
  4. Results & Analysis
    • Detailed performance metrics
    • Test results across categories
    • Comparative analysis with baselines
    • Error analysis and mitigation strategies

More information on this theme will be revealed on 20th November before EOD

 

UX Design Problem Statement – Designing the Future of Fashion Trend Discovery

Overview:

This theme invites participants to reimagine how designers, planners, and buyers discover and track fashion trends. Current tools are often static, complex, and uninspiring. The goal is to design a B2B product that is engaging, intuitive, and personalized, transforming how users monitor market trends, competitors, and new inspirations.

Challenge:

Design a product tailored for personas like designers, planners, and merchandisers. The solution should balance data insights with creative freedom, solving inefficiencies in current trend discovery processes. It must address key jobs, such as tracking trends, analyzing competitors, and planning optimal assortments.

Key Considerations:

  • User-Centered Design: Align with the needs of designers, planners, and buyers.
  • Seamless UX: Make the product interactive, intuitive, and easy to navigate.
  • Future-Ready: Ensure adaptability to evolving trends and industry needs.
  • B2B Focus: Provide scalable, professional functionality for retail workflows.

More information on this theme will be revealed on 20th November before EOD

 

Prizes

Main Prizes
1st Prize
INR 1,00,000
2nd Prize
INR 65,000
3rd Prize
INR 35,000
starts on:
Nov 13, 2024, 12:30 PM UTC (UTC)
closes on:
Dec 15, 2024, 06:29 PM UTC (UTC)

Social Share

Help & Support

Please contact event admin
Jyothishree S at jyothishree@hackerearth.com
Notifications
View All Notifications

?