Features
- Friendly and safe Natural Language Conversation
- Non-biased and non-toxic content
- Real Time highly accurate Speech Recognition in 80 languages
- Real Time Text to Speech using Neural realistic and natural voices in 125 languages
- Custom AI Large Language Model trained using alignment techniques such as supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF)
- Drag-and-Dropt .pdf and docs with custom data files to train the AI
- Option to use OpenAI Davinci, GPT, and ChatGPT AI models in parallel
- Web Search, in addition to AI
- Option to use custom trained LLMs
- Integration with AUTOMATIC1111 / Stable Diffusion Webui
- Emotional Support AI
- Create your custom Avatar using Midjourney and Stable Diffusion and use Denise's Avatar Editor to make it talk
- Customize the AI Personality traits
- Integration with oobabooga text generation webui
- Short and long term memory
- Ultra-realistic Avatars to speak aloud the output in real-time - Ver 1.0
- Can work in tandem with Denise Legacy to use legacy CG Avatars
- Generate AI Images during the conversation
- Option to add custom data from documents and .pdf files
- Avatar Editor
- API
- Open Source whole project code
- 60 days full access to DaVinci Infinito, a text, blog, voice, image, Content Rewriter and Summarize AI Service *
- Available soon
New Photo realistic Digital Humans Avatars
Current CG Avatar's Improvements
Create your custom Talking Avatar for Deniise 2.0 with
Midjourney, Stable Diffusion, Leonardo, DALLE, etc.
The images below characterize the new Deniise 2.0 Avatar and don’t necessarily represent all available avatars.
Milestone #1
Neural Text to Speech Voices
Custom AI (LLM) training
OpenAI GPT3 and GPT4 integration
Servers
Creation of the first Human-Like Character
Authentication System
Milestone #2
Finish Interface Design
Installation system
Web Site
User Guide
Community Blog
Help and Support Systems
Release Denise 2.0 Beta
Milestone #3
Real-Time Lip-syncing - Version 2
Conection with LLMs based on Alpaca and Open Assistant
API
FAQ - Frequently Asked Questions
Deniise 2.0 is a user-friendly desktop (the mobile version is on its way!) application designed to change how you interact with advanced Artificial Intelligence. Different from the original Denise whose main function was being a Virtual Assistant, with Deniise 2.0, you can engage in natural, voice-based conversations with an AI Avatar that feels like a virtual companion and emotional support coach. The goal with this new version is to offer a friendly, always-available companion who will listen to you, remember your previous conversation, and talk to you as a good friend and coach would do. Virtual Assistant features will also be present to help you with your daily activities and operating your computer and mobile device. By using the new advances in synthetic voices, Denisse’s responses are brought to life with incredibly realistic voices through the power of new Neural Text to Speech technology.
Deniise 2.0 builds upon the success of its predecessors, Virtual Assistant Denise (2005) and Denise Legacy (2022).
Deniise 2.0 is the culmination of years of development and innovation. This latest version works seamlessly on your desktop and will soon be available on mobile devices, giving you the freedom to connect with the AI Avatar wherever you go.
This latest version can be used independently or in conjunction with Denise Legacy. Deniise 2.0 harnesses the power of a custom-trained LLM (large language model) and OpenAI’s Davinci, GPT3, GPT4, and ChatGPT4 AI models when used separately. Alternatively, users can integrate Deniise 2.0 with Denise Legacy, where responses can be relayed to Denise Legacy, allowing her to vocalize them using her legacy Avatars.
One notable enhancement in Deniise 2.0 is the improved real-time lip-syncing engine, which seamlessly animates ultra-realistic Avatars during conversations, adding a remarkable level of realism. However, Deniise 2.0 is only the initial step in our more ambitious undertaking, Deniise 2.0+. This upcoming Kickstarter campaign will introduce a more robust AI software that will take the capabilities of Deniise 2.0 to new heights. To support this endeavor, we will launch a Kickstarter campaign soon, and we can’t wait to share this next chapter with you.
Stay tuned for updates and be part of the future of AI interaction with Deniise 2.0. Whether you use it independently or alongside Denise Legacy, this application is set to revolutionize how we engage with AI technology. Get ready to experience the next level of conversational AI!
We are thrilled to announce that current Denise 1.9 and Denise Legacy customers will receive discounts on Deniise 2.0 purchases. As for new customers, we offer a lifetime affordable fee to support our ongoing efforts. At this stage, the core code of Deniise 2.0 will not be accessible until we achieve a safe and stable AI Model. Our ultimate goal is to transform this project into an Open Source initiative, and we are actively working towards that. Excitingly, we are set to begin the Beta test phase in July.
Davinci Infinito is an advanced web platform that empowers users to leverage state-of-the-art Artificial Intelligence technology to generate diverse text content forms, including articles, blogs, ads, and media in 37 languages. The platform offers limitless possibilities in terms of what users can create. Additionally, users can effortlessly generate AI images by simply describing the desired image using OpenAI’s DALL-E-2 and Stable Diffusion (provided by Stability.ai). Davinci Infinito is supported by various powerful OpenAI models, including GPT4, GPT3 Turbo, GPT3: Davinci, Curie, Babbage, and Ada, alongside custom-trained AI models. Furthermore, the platform provides an AI Code feature, enabling users to generate code in any programming language with the assistance of AI.
Infinito also encompasses an interactive AI Chat system, allowing users to pose queries and receive tailored results. The platform includes 42 AI Chat Bot Assistants, specializing in specific areas, similar to the ChatGPT platform. Additionally, users can transcribe audio and video files using the Speech to Text feature, utilizing OpenAI’s Whisper model.
In addition to the features, as mentioned earlier, users can produce AI voiceovers using over 540 voices spanning 140 languages and dialects. They can combine up to 20 voices in a single task and synthesize text with a maximum length of 100,000 characters, all powered by Neural text-to-speech voices. The writing assistant capability of Davinci Infinito seamlessly integrates into any business context, providing significant time-saving benefits.
Visit DaVinci Infinito!