This project is designed to empower individuals who are preparing for their first impression, public speech, or even everyday social interactions. The platform allows users to record a video of themselves—be it a self-introduction, an approach, or a presentation—and then processes the video to extract valuable insights.
PitchPerfect leverages three cutting-edge AI technologies to enhance user experience:
-
Gemini AI: Provides intelligent feedback on speech clarity, body language, and delivery.
-
DeepFace: Analyzes facial expressions in real time to assess emotional reactions.
-
Whisper AI: Transcribes the recorded video into text for review and refinement.
Video Recording & Transcription: Users can easily record their video using their webcam. Once the recording is stopped, the platform automatically converts the video into text, allowing users to review and refine their speech or introduction.
AI-Powered Feedback: Leveraging the advanced capabilities of Gemini AI, the platform analyzes the recorded video to generate personalized feedback. This feedback includes tips on improving speech clarity, body language, and overall delivery to help users present their best self.
Facial Expression Analysis: Using DeepFace technology, the live camera continuously monitors and detects facial expressions during the recording. This feature visualizes the user's emotional reactions in real time, providing insights into how their expressions might be perceived by an interviewer or audience.
Interactive User Interface: The platform offers a modern, responsive interface built with Tailwind CSS. Custom colors create a sleek and professional look. Users benefit from a typewriter-style welcome animation, clear call-to-action buttons, and intuitive controls, such as a dedicated retry option to restart the recording process.
Whether you're gearing up for an important interview, preparing for a speech, or simply looking to boost your social confidence, this platform provides a comprehensive toolset to help you hone your presentation skills and make a lasting impression.
Video Recording: Capture your self-introduction or interaction using your webcam.
Transcription: Convert the recorded video into text automatically.
AI-Powered Feedback: Generate feedback on your performance using Gemini AI.
Facial Expression Detection: Leverage DeepFace to detect and visualize your facial reactions during recording.
Typewriter Animation: Engaging welcome messages to guide users as they enter the platform.
Responsive UI: Built with Tailwind CSS, featuring custom UI colors for a modern look.
Retry Option: Easily reset your recording session with a dedicated retry button.
Clone the repository: git clone https://github.com/R-umaria/HackCanada.git
How to use:
- Click 'Start Recording' to begin capturing your video. Your video is automatically converted to text after stopping the recording.
2.Feedback: Click Get Feedback to view AI-generated feedback on your performance.
-
Facial Expression Visualization: The live camera uses DeepFace to detect and display your facial reactions during the recording.
-
Retry: Use the Retry button to clear the current session and start over.
Frontend: HTML, Tailwind CSS, and JavaScript Webcam & Media Recording: Browser's getUserMedia and MediaRecorder APIs AI & Feedback: Gemini AI for feedback generation Facial Expression Analysis: DeepFace for face detection and reaction visualization Transcription: (Placeholder) Video-to-text conversion functionality
Click the image above or watch the tutorial here to learn how to use PitchPerfect effectively.



