虛擬人範例 https://youtu.be/nJd_2mJ4u-I?t=91
【Responsibilities】
1. 研究並測試 市面上主流圖像/影像型AIGC 工具與模型(如 Stable Diffusion、DALL·E、MidJourney、Runway ML、Pika Labs、Sora 等),並分析其 生成品質、運行成本、適用性與技術限制。
2. 開發 API 串接與整合方案,將 AIGC 生成技術無縫應用於內部平台或客戶專案,確保效能與穩定性。
3. 研究 Prompt Engineering、ControlNet、LoRA 微調 等技術,以優化 AIGC 內容生成效果,並適配不同應用場景(如遊戲、動畫、直播虛擬人)。
4. 參與 AIGC 模型微調與訓練,研究 LoRA、DreamBooth、Fine-tuning、Diffusion Models 強化,提升生成內容品質。
5. 探索 AIGC 技術在虛擬人領域的創新應用,包括表情生成、動作合成等。
6. 探索多模態生成技術(如 CLIP、LLaVA、音訊轉影片技術),並評估其應用場景。
7. 關注 AIGC 技術發展趨勢,定期分析業界競品,提供技術選型與應用建議。
8. 整合 提示式檢索增強生成技術 (RAG)
9. 整理測試報告、技術文檔,確保團隊成員能夠理解並使用相關技術與工具。
1. Research and test mainstream AIGC tools and models on the market (such as Stable Diffusion, DALL·E, MidJourney, Runway ML, Pika Labs, Sora, etc.), analyzing their generation quality, operational costs, applicability, and technical limitations.
2. Develop API integration and implementation solutions, ensuring seamless application of AIGC generation technology within internal platforms or client projects while maintaining performance and stability.
3. Study Prompt Engineering, ControlNet, LoRA fine-tuning, and other techniques to optimize AIGC content generation effects, adapting them to different application scenarios (e.g., gaming, animation, live-streaming virtual humans).
4. Participate in AIGC model fine-tuning and training, researching LoRA, DreamBooth, fine-tuning, and Diffusion Model enhancements to improve content generation quality.
5. Explore innovative applications of AIGC in the virtual human field, including facial expression generation and motion synthesis.
6. Investigate multi-modal generation technologies (such as CLIP, LLaVA, and audio-to-video technologies) and evaluate their application scenarios.
7. Keep track of AIGC technology trends, regularly analyze industry competitors, and provide technical recommendations for model selection and application.
8. RAG
9. Compile test reports and technical documentation to ensure team members understand and can utilize relevant technologies and tools.
【Requirements】
1. 碩士以上學歷,優先考慮具備資工背景的候選人。
2. 精通Python,具備5年以上開發經驗。
3. 熟悉 生成式 AI(Generative AI),如 Diffusion Models、GAN、Transformer-based 生成技術等,並能理解其原理與應用場景。
4. 了解視覺內容處理技術,如 OpenCV、FFmpeg、Blender API、Unreal Engine API、Stable Diffusion WebUI 等。
5. 具備 API 整合與自動化腳本開發能力,能夠串接第三方 AI 生成服務並優化其應用流程。
6. 具有電腦圖學computer vision與AI影像處理等相關經驗。
1. Master's degree or above, with preference for candidates with a background in computer science.
2. Proficient in Python, with 5+ years of development experience.
3. Familiar with Generative AI technologies, including Diffusion Models, GANs, and Transformer-based generation techniques, with a strong understanding of their principles and application scenarios.
4. Knowledge of visual content processing technologies, such as OpenCV, FFmpeg, Blender API, Unreal Engine API, and Stable Diffusion WebUI.
5. Experience in API integration and automation scripting, with the ability to connect and optimize third-party AI generation services.
6. Experience in computer graphics and AI-based image processing.