AI Research 技术调研

标签: rope

此标签下有7条笔记。

  • 2026年7月16日

    3D Diffuser Actor: Policy Diffusion with 3D Scene Representations

    • manipulation
    • diffusion-policy
    • 3d-scene-representation
    • relative-attention
    • rope
    • keypose
    • rlbench
    • calvin
    • imitation-learning
    • embodied-ai
  • 2026年6月25日

    LaVie: High-Quality Video Generation with Cascaded Latent Diffusion Models

    • t2v
    • video-diffusion
    • latent-diffusion
    • cascaded
    • temporal-attention
    • rope
    • joint-image-video
    • vimeo25m
    • open-source
  • 2026年6月25日

    Hunyuan-DiT: A Powerful Multi-Resolution Diffusion Transformer with Fine-Grained Chinese Understanding

    • t2i
    • diffusion-transformer
    • dit
    • chinese
    • bilingual
    • multi-resolution
    • rope
    • recaptioning
    • mllm
    • open-source
  • 2026年6月25日

    Lumina-Next: Making Lumina-T2X Stronger and Faster with Next-DiT

    • text-to-image
    • diffusion-transformer
    • next-dit
    • flow-matching
    • rectified-flow
    • rope
    • resolution-extrapolation
    • few-step-sampling
    • multilingual
    • unified-generation
  • 2026年6月25日

    Lumina-T2X: Transforming Text into Any Modality, Resolution, and Duration via Flow-based Large Diffusion Transformers

    • dit
    • flow-matching
    • flag-dit
    • text-to-image
    • text-to-video
    • text-to-3d
    • text-to-speech
    • rope
    • resolution-extrapolation
    • unified-generation
  • 2026年6月25日

    OminiControl: Minimal and Universal Control for Diffusion Transformer

    • dit
    • flux
    • controlnet
    • subject-driven
    • image-conditioning
    • lora
    • rope
    • dataset
  • 2026年6月25日

    UNO: Less-to-More Generalization — Unlocking More Controllability by In-Context Generation

    • subject-driven
    • customization
    • in-context-generation
    • dit
    • flux
    • lora
    • rope
    • multi-subject
    • data-synthesis

  • GitHub