1.具有硕士或博士学位，有人工智能领域的项目经历。
2.具备基本的机器学习和深度学习知识，了解常见的机器学习和深度学习模型，如Transformer，CNN，Seq2Seq等，能熟练使用PyTorch、Transformers等深度学习框架
3.自驱力强，乐于探索新技术，积极跟团队成员沟通讨论、合作。
4.加分项：
○ 在人工智能领域的会议或期刊发表过论文， 如CVPR、ICCV、ECCV、NeurIPS、ICML、ICLR、ACL、EMNLP等。
○ 有大模型训练、微调经验，能熟练使用PEFT
○ 有网站或AI应用设计和开发经验，参与的AI开源项目github stars数100+
○ 有较强的PPT制作、Demo视频剪辑能力
机器学习实验室重点研究方向：
1. Vision-Language Modeling (VLM): advancing 2D and 3D vision-language tasks via innovating model architectures, training algorithms, and collecting high-quality data, through which unifying various tasks in one model. Our recent focus: 3D-VL; VLM for long-form and streaming video understanding; video and 3D generation.
References:
● 3D-VisTA https://3d-vista.github.io/
● UltraEdit https://ultra-editing.github.io/
2. Vision-Language-Action (VLA): integrating vision and language with action-oriented tasks such as planning& control in simulated engine and real robotics. Our recent focus: dexterous manipulation; mobile manipulation.
References:
● LEO https://embodied-generalist.github.io/
● JARVIS-1, OmniJARVIS,https://omnijarvis.github.io/
3. AGI Agents: exploring the development of AI agents capable of general tool-use, problem solving, feedback reflection, and continual learning. Our recent focus: multi-modal agents; agent tuning; reflecting and learning from user feedback.
References:
● CLOVA https://clova-tool.github.io/
● VideoAgent https://videoagent.github.io/
允许3个月内投递2个职位，请选择适合的职位进行投递
立即投递
