Klug.pptx
Download as PPTX, PDF0 likes21 views
The document describes a proposed virtual trial room called Klug Reflector that uses AI, machine learning, and deep learning to provide an efficient and personalized virtual clothing try-on experience. It aims to streamline the traditional time-consuming clothing trial process. Key features include gesture recognition for user interaction, integration of computer vision technologies, and increased sales potential through personalized recommendations. The system architecture incorporates modules for user selection, body detection, clothing search, video masking, and online payment. Experimental results demonstrate the system's capabilities and comparisons highlight its advantages over existing solutions.
![References
[1] Ari Kusumaningsih; Arik Kurniawati; Cucun Very Angkoso,
Eko Mulyanto Yuniarno; Mochammad Hariadi,
[2] Srinivasan K.,Vivek S., "Implementation of virtual fitting
room using image processing"
[3] Md. Farhan hamid, md. Ashraful alam, "virtual wardrobe
for physically impaired using Microsoft kine ct sensor"
[4] Lianghua Huang, Di Chen, Yu Liu, Yujun Shen, Deli Zhao,
Jingren Zhou||Composer: Creative and Controllable Image
Synthesis with Composable Conditions
[5] Miaolong Yuan; Ishtiaq Rasool Khan; Farzam Farbiz; Susu
Yao;||A Mixed Reality Virtual Clothes Try-On System](https://image.slidesharecdn.com/klug-240201050959-997aaf3f/85/Klug-pptx-14-320.jpg)
More Related Content
Similar to Klug.pptx (20)
More from Sharad Institute of Technology,college of Engineering,Yadrav (6)
Recently uploaded (20)
Klug.pptx
- 1. KLUG REFLECTOR USING AIML & DL FOR THE FASHION INDUSTRY Dr.S.S.Gurav Departemnt Electronics & Computer Sharad Institute of Technology, College of Engineering,Yadrav
- 2. Introduction Time Efficiency: Traditional clothing trials are time- consuming; the virtual trial room streamlines the process. Online vs. Offline: Bridges the gap between online and offline shopping, catering to diverse consumer preferences. Gesture Recognition: Identifies face and body gestures, enhancing user interaction and personalization. Technological Integration: Utilizes Open CV, Python, Visual Studio, and AI technologies for seamless control and user experience. Sales Boost: Forbes Insights suggests increased personalization through virtual trial rooms positively impacts sales and profits.
- 3. Background of the Paper Time Efficiency: Klug Reflector streamlines the clothing trial process, reducing time constraints in physical and online stores through its virtual trial room. Fabric Quality Assurance: By incorporating AIML & DL, Klug Reflector ensures fabric quality preservation, offering customers a reliable and satisfactory trial experience. Diverse Options: Overcoming limitations of traditional methods, the virtual trial room provides a wide array of clothing choices, enriching the overall shopping experience. Hygienic Shopping: Klug Reflector eliminates the impracticality of sanitizing all clothes by introducing a contact-free trial, ensuring a hygienic shopping environment amid COVID-19.
- 4. Related works Inclusive Clothing Access: Md. Farhan Hamid's research focuses on aiding physically impaired individuals, using sensors to track head and eye movements for clothes selection, prioritizing accessibility for the disabled. Regional Limitations: Ari Kusumaningsih's work explores a virtual trial room for Madura Batik clothing, emphasizing online trials. However, the drawback lies in limited offerings for international customers, primarily catering to local use. Online Shopping Platform: Srinivasan K.'s method, utilizing a fixed webcam and MATLAB, caters to online shopping, lacking offline applications and precise depth readings due to the absence of a Kinect sensor.
- 6. A virtual trial room where a person can try the model they require with the precise measurements and color they desire together with the change in the background and environment they are entering will be one of the ideal solutions provided by the proposed system. They will spend less time traveling and changing because this requires no physical touch and offers a variety of customization possibilities that are often not available in the current system. Consumers can visit any store in the world and try the clothes on to see if they will fit or not, which helps them avoid restricting themselves to a single store or location.Using ML and DL we can filtering and makeup video stream. It look more beautiful.
- 7. User Module: Enables clothing selection, color choice, and style preferences through a user-friendly interface. Actions like adding to a cart or marking as a favorite can be performed seamlessly using AI, media-pipe, and hand motions. Detection Module: Utilizes Media-pipe structure for body point detection, enabling the masking of the video stream for a specific system. Searching Module: Employs image processing and deep learning with a virtual mouse system, allowing customers to navigate and find clothes on the website. Human hand estimation and gestures enhance system interaction. Masking Module: Disguises the user's clothing based on identified body joints. Filtering Module: Enhances the video stream aesthetics using ML & DL. Output Module: Provides real-time masked video output. Booking Module: Facilitates online payment for chosen clothes.
- 8. Results Integral Image Fig. Integral image Rectangular features, representing pixel sum differences, are computed in various configurations. Convoluting these regions through the image produces filters for Ada Boost. The proposed integral image method significantly reduces the typical lengthy computing time.
- 9. Haar like Features Haar-like feature User ditection Face recognition
- 10. Facial Landmarks Detection Output
- 11. Discussion The Klug Reflector, integrating AIML & DL in fashion, presents a transformative solution. Addressing time constraints, fabric quality, and COVID-19 concerns, the virtual trial room enhances offline shopping productivity. The system's innovative use of ML and DL offers a personalized, efficient, and infection-free shopping experience. Comparison with existing studies underscores its uniqueness. Two notable case studies, focusing on physically impaired individuals and virtual trial rooms for regional attire, validate the system's versatility. The proposed system not only streamlines traditional clothing trials but also aligns with global trends, promising significant benefits to the fashion industry.
- 12. Conclusions The virtual changing room app operates with just a front image, effectively using 2D graphics for versatile user interactions. The matching technique performs well in diverse scenarios, especially in standard postures, demonstrating satisfactory experiment performance. Various model fitting options exist, including homographic transformation for simultaneous joint matching, though it requires more processing. Realistic video streams can be generated by combining images from different perspectives. Rendering 3D models based on current angles, positions, and utilizing a physics engine enhances the virtual fitting experience.
- 13. Future work To calculate size of a person. Filtering a face of a person. Update website. Make 3D version of over project. Add booking system. Add online payment option. Make it as a app for mobile user.
- 14. References [1] Ari Kusumaningsih; Arik Kurniawati; Cucun Very Angkoso, Eko Mulyanto Yuniarno; Mochammad Hariadi, [2] Srinivasan K.,Vivek S., "Implementation of virtual fitting room using image processing" [3] Md. Farhan hamid, md. Ashraful alam, "virtual wardrobe for physically impaired using Microsoft kine ct sensor" [4] Lianghua Huang, Di Chen, Yu Liu, Yujun Shen, Deli Zhao, Jingren Zhou||Composer: Creative and Controllable Image Synthesis with Composable Conditions [5] Miaolong Yuan; Ishtiaq Rasool Khan; Farzam Farbiz; Susu Yao;||A Mixed Reality Virtual Clothes Try-On System