AI Twin Avatar: Training Video Best Practices
Ready to create an AI Avatar Twin that sounds, looks and moves like you? First, you need to create a training video that the AI Model will use to create your digital clone. The better the training video- the more realistic your AI avatar will be.
To learn more about how to create an AI Avatar, see this article
Record in HD: Quality is key! An iPhone 12 or above is enough.
Framing: When framing your shot, make sure to level the camera to your eye line, and fill up at least 2/3 of the frame, leaving some space at the top of the frame.
Sit Still: Try to maintain a steady posture and avoid unnecessary movements. Sitting still allows the AI to better capture your facial expressions and ensure a more accurate avatar.
Pause Between Sentences: After finishing each sentence, take a short pause and close your lips. This helps the AI recognize natural pauses in speech and improves lip-syncing accuracy.
Even Lighting on the Face: Ensure that your face is evenly lit. Balanced lighting helps the AI capture facial details more effectively, reducing shadows that might distort features.
Increase Energy and Emotion: Speak with about 25% more energy and emotion than you normally would. The AI smooths out speech, so adding extra emphasis ensures the final result doesn’t sound flat or monotone.
Hand Movements and Gestures: To make your AI Twin copy hand movements and gestures, include them in your video. The AI will copy your movements and add them at relevant parts of your video, making your avatar look super realistic. Make sure that you hands stay in the frame and are not cut off for the entirety of the training video.
Strong Shadows: Avoid having harsh shadows on your face, as they can interfere with the AI’s ability to detect facial features accurately.
Sunglasses or Headphones: Do not wear sunglasses or headphones while recording. These items can obscure your face and create unnatural obstacles for the AI’s recognition process.
Background Distractions: Avoid background activities like pets, moving objects, or TV screens. These distractions can confuse the AI during processing.
Paintings or Photos with Faces: Ensure there are no portraits, paintings, or pictures with faces in the background. These can interfere with the AI's ability to focus on your face.
Blurry Camera: Make sure your camera is in focus. Blurry footage will negatively affect the AI’s ability to map your facial features.
Avoid Not Looking at the Camera: Speak directly into the camera to ensure the AI accurately captures your eye contact and facial engagement with your audience.
Hair Across the Face: Make sure your face is fully visible, with no hair obscuring it. Hair across your face can interfere with the AI’s ability to analyze your features correctly.
Ready to submit your training video and create your AI Avatar? Read this article
To learn more about how to create an AI Avatar, see this article
Do’s for Training Video Recording:
Record in HD: Quality is key! An iPhone 12 or above is enough.
Framing: When framing your shot, make sure to level the camera to your eye line, and fill up at least 2/3 of the frame, leaving some space at the top of the frame.
Sit Still: Try to maintain a steady posture and avoid unnecessary movements. Sitting still allows the AI to better capture your facial expressions and ensure a more accurate avatar.
Pause Between Sentences: After finishing each sentence, take a short pause and close your lips. This helps the AI recognize natural pauses in speech and improves lip-syncing accuracy.
Even Lighting on the Face: Ensure that your face is evenly lit. Balanced lighting helps the AI capture facial details more effectively, reducing shadows that might distort features.
Increase Energy and Emotion: Speak with about 25% more energy and emotion than you normally would. The AI smooths out speech, so adding extra emphasis ensures the final result doesn’t sound flat or monotone.
Hand Movements and Gestures: To make your AI Twin copy hand movements and gestures, include them in your video. The AI will copy your movements and add them at relevant parts of your video, making your avatar look super realistic. Make sure that you hands stay in the frame and are not cut off for the entirety of the training video.
Dont’s for Training Video Recording: Training Video Recording:
Strong Shadows: Avoid having harsh shadows on your face, as they can interfere with the AI’s ability to detect facial features accurately.
Sunglasses or Headphones: Do not wear sunglasses or headphones while recording. These items can obscure your face and create unnatural obstacles for the AI’s recognition process.
Background Distractions: Avoid background activities like pets, moving objects, or TV screens. These distractions can confuse the AI during processing.
Paintings or Photos with Faces: Ensure there are no portraits, paintings, or pictures with faces in the background. These can interfere with the AI's ability to focus on your face.
Blurry Camera: Make sure your camera is in focus. Blurry footage will negatively affect the AI’s ability to map your facial features.
Avoid Not Looking at the Camera: Speak directly into the camera to ensure the AI accurately captures your eye contact and facial engagement with your audience.
Hair Across the Face: Make sure your face is fully visible, with no hair obscuring it. Hair across your face can interfere with the AI’s ability to analyze your features correctly.
Ready to submit your training video and create your AI Avatar? Read this article
Updated on: 29/09/2024
Thank you!