Responsibilities • Develop and optimize image generation pipelines for text-to-image, image-to-image, image editing, inpainting, and style transfer tasks. • Research, reproduce, and improve generative models based on diffusion models, flow matching, rectified flow, or other image generation frameworks. • Fine-tune and adapt image generation models for specific styles, characters, products, or application scenarios. • Work with techniques such as LoRA, ControlNet, IP-Adapter, DreamBooth, prompt engineering, and reference-based generation. • Build and maintain training, inference, and evaluation pipelines using PyTorch and related tools. • Improve generation quality, controllability, inference speed, and GPU memory efficiency. • Prepare, clean, and organize image-text datasets for model training and evaluation. • Analyze experiment results and write clear technical documentation. Requirements • Currently enrolled as a Bachelor’s, Master’s, or Ph.D. student in Computer Science, Artificial Intelligence, Computer Vision, Electrical Engineering, or a related field. • Strong programming skills in Python and hands-on experience with PyTorch. • Solid understanding of deep learning, computer vision, and generative models. • Familiar with diffusion-based image generation models such as Stable Diffusion, SDXL, Flux, DiT, or similar models. • Experience with at least one of the following is preferred: LoRA, ControlNet, IP-Adapter, DreamBooth, Hugging Face Diffusers, ComfyUI, or Automatic1111. • Ability to read and reproduce research papers in generative AI or computer vision. • Experience with model fine-tuning, dataset preparation, or inference optimization is a plus. • Familiarity with Linux, Git, Docker, CUDA, TensorRT, or ONNX is a plus. • Strong self-learning ability, good communication skills, and interest in building practical AIGC applications. • Good communication skills in English