Prepare a structural reference from a photo or an extracted video frame: depth for near/far layout, body pose for a person’s posture, or line art for visible contours. Keep the source and result together.
Explore the tools · 了解工具Prepare a structural reference from a photo or an extracted video frame: depth for near/far layout, body pose for a person’s posture, or line art for visible contours. Keep the source and result together.
Open an image or send a selected frame here → choose Depth map, Pose map or Line art → generate and review. Save the result for a creative workflow that accepts that reference type.
Depth is an estimated relative map, not measured distance or a 3D model. Pose uses RTMW-X for up to four people and draws body pose only, without detailed hand or face landmarks. Depth and pose download models on first use. Results and speed depend on the image, device and browser.
They are image files you can try in a workflow that accepts depth, body pose or line-art inputs. Check that workflow’s required map format, orientation and model. There is no direct generator integration or guarantee that every control model accepts the same map.
Single-file tools are free, with no watermark. Media is processed in your browser, not uploaded. Sign in for 3 shared Pro batch trials, up to 3 inputs per batch; a trial counts when export starts successfully. Pricing & availability · Privacy · Operator & terms · support@murmurlayer.com
从照片或视频抽帧中准备结构参考：深度图表达空间远近，骨架图提取人体姿态，线稿保留可见轮廓。原图与结果保留在同一工作区。
打开图片或从素材选集传入一帧 → 选择深度图、骨架图或线稿 → 生成并检查。在支持该参考类型的创作流程中使用保存的结果。
深度是相对远近估计，不是测距或三维模型。骨架使用 RTMW-X，最多识别 4 人，仅绘制身体姿态，不含精细手部或面部关键点。深度与骨架首次使用需下载模型；效果及速度受图片、设备和浏览器影响。
可在接收深度、身体姿态或线稿图片的流程中尝试。请核对目标流程要求的图像格式、方向与控制模型。本站未直接接入生成器，也不保证所有控制模型都接受相同参考图。
单文件工具免费、无水印，素材在本机浏览器处理，不上传。登录后有 3 次共用 Pro 批量试用，每批最多 3 个输入；成功发起导出才计次。 价格与开通状态 · 隐私 · 运营信息与条款 · support@murmurlayer.com