Heygem is the Silicon Intelligence launch of theOpen SourceDigital Humanmodel, designed for Windows systems. Based on the advanced A! technology, it takes only 1 second of video or 1 photo to complete the image and voice cloning of a digital human in 30 seconds, and synthesize a 4K UHD video in 60 seconds. Heygem supports multi-language output, multi-expression actions, and 100% mouth matching capability, which maintains a highly realistic effect in complex light and shadow or occlusion scenarios. Heygem is based on full offline operation mode, protects user privacy, supports low configuration hardware deployment, greatly reduces the threshold of use, and provides efficient and low-cost digital human solutions for content creation, live broadcasting, education and other scenarios.
Heygem Features
- Seconds cloning: only 1 second video or 1 photo is needed to complete the digital person image and voice cloning, 30 seconds to complete the cloning, 60 seconds to synthesize 4K UHD digital person video.
- Efficient inference: inference speed up to 1:0.5, video rendering compositing speed up to 1:2.
- High-quality output: Supports 4K Ultra HD, 32 fps video output, surpassing the 24 fps standard for Hollywood movies.
- Multi-language support: The cloned digital man supports 8 languages output to meet the global market demand.
- Unlimited Cloning: Support unlimited cloning of digital human images and voices, unlimited synthesized videos.
- 100% Mouthpiece Matching: Highly realistic mouthpiece matching in complex lighting, occlusion or side angles
- Low-configuration can run: support Docker one-click deployment, the minimum only need NVIDIA 1080Ti graphics card to run.
Official website link:http://heygem.ai