progress 1 Group 13 Project presentation final year

Group 13 Project presentation final year

1. SHRI SAI SHIKSHAN SANSTHA’S NAGPUR INSTITUTE OF TECHNOLOGY, NAGPUR Department of Information Technology 2025-2026 Session 2024-2025 Project Progress Seminar Topic : Occasion Based Clothing Recommender System using AI Project Members 1) Ayush Gaikwad 2) Ganesh Awghade 3) Harsh Kadu 4)Chaitanya Nasare 5)Abhijeet Dhuppe @ NIT,Nagpur Guided By Prof. Dhananjay Kondekar

2. CONTENT Introduction Methodology Problem Identification Proposed Solution Implementation of UI Probable Outcomes References NIT,Nagpur

3. INTRODUCTION NIT,Nagpur In today’s fast-paced world, choosing the right outfit for the right occasion is often a confusing and time-consuming task. People struggle to decide what to wear for different events such as weddings, interviews, festivals, parties, or casual outings. The growing variety of fashion trends and personal preferences further adds to this complexity. An Occasion-Based Clothes Recommendation System using Artificial Intelligence (AI) offers a smart solution to this problem. By analyzing the occasion, user profile (age, gender, body type, preferences, budget), and contextual factors (weather, season, latest trends), the system can recommend the most suitable clothing options.

4. start Problem Understanding & Requirement Data Collection & Preprocessing System Design ( Frontend,Backend ) Model Training & Evaluation Output Generation (Outfit Suggestion) Deployment & Testing

5. Methodology The methodology of this project involves a sequence of steps to design, develop, and evaluate an AI-based clothes recommendation system. 1. Problem Understanding & Requirement AnalysisIdentify the challenges in outfit selection for different occasions.Define system requirements: inputs (occasion, user profile, wardrobe data, weather), outputs (recommended outfits). 2. Data Collection & PreprocessingDatasets : Fashion images, outfit datasets (e.g., DeepFashion , Polyvore, ModaNet ).Attributes: Clothing type, color , fabric, style, season, occasion tags.Preprocessing:Image resizing, normalization.Annotating clothes with attributes (e.g., “formal shirt”, “wedding saree”).Removing noise and duplicates. 3. Feature Extraction Use Computer Vision (CNN models like ResNet , VGG, EfficientNet ) to extract visual features ( color , texture, pattern).Extract textual features (description, tags, occasion labels) using NLP techniques (TF-IDF, BERT).Combine features to build a multi-modal representation of clothing items. NIT,Nagpur

6. PROBLEM IDENTIFICATION NIT,Nagpur Selecting appropriate clothing for different occasions is often a confusing and time-consuming task. With the wide variety of fashion choices available, people face challenges in deciding what to wear for events such as weddings, interviews, parties, festivals, or casual outings.Factors like personal preferences, body type, budget, weather conditions, and rapidly changing fashion trends make the decision even more difficult. As a result, individuals may end up wearing unsuitable outfits, leading to discomfort, lack of confidence, or poor impressions in social and professional settings. There is a clear need for a smart, AI-driven recommendation system that can reduce decision-making stress by suggesting the most suitable clothing based on the occasion and user context

7. PROPOSED SOLUTION NIT,Nagpur Build an AI-driven system that:1. Understands the occasion (wedding, office meeting, sports event, casual outing, festival, etc.) Analyzes user profile (age, gender, body type, fashion preferences, budget). Considers external factors (season, current fashion trends, location, weather). Recommends clothing styles (formal wear, ethnic, casual, sportswear, party wear, etc.) with images and possible online purchase links.

8. Implementation of UI NIT,Nagpur

9. PROJECT OUTCOME NIT,Nagpur Outcome: The developed system will provide personalized clothing recommendations based on the user’s occasion, preferences, and contextual factors such as season and fashion trends . Users will receive smart outfit suggestions for events like weddings, interviews, parties, festivals, or casual outings, along with accessories and matching styles. The outcome will be: A user-friendly platform (web/app) that reduces confusion in outfit selection. Increased confidence and satisfaction by ensuring users wear the right outfit for the right occasion. Integration with e-commerce platforms for convenient outfit purchasing. A foundation for future enhancements such as wardrobe scanning, AI stylist chatbots, and trend-based updates.

10. REFERENCES 1. A Review of Modern Fashion Recommender Systems — Yashar Deldjoo et al. (2022) — survey. PDF: https://arxiv.org/pdf/2202.02757 2. Study of AI-Driven Fashion Recommender Systems — Shaghayegh Shirkhani , Hamam Mokayed , Rajkumar Saini, Hum Yan Chai (2023) 3. A Comprehensive Analysis of Outfit Recommendation — Gursimran Kaur, Hrithik Malhotra, Tanmaya Gupta (2021) — survey on content-based, collaborative, hybrid, ontology-based approaches. PDF: https://ijcaonline.org/archives/volume183/number12/kaur-2021-ijca-921413.pdf 4. Skin-Tone and Occasion Oriented Outfit Recommendation System — Digant Garude et al. (2019) — focusing on skin tone + occasion etc. SSRN link: https://ssrn.com/abstract=3368058 NIT,Nagpur

11. Thank You