OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

3D Camera Control · Multi-Angle Image Generation

3D Camera Control · Multi-Angle Image Generation is a bilingual Python and Gradio application for creating image edits from chosen camera viewpoints. It combines a Three.js camera widget and sliders for azimuth, elevation and distance with an image-generation backend.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

3D Camera Control · Multi-Angle Image Generation — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when a still image needs alternate viewpoints and you want to specify the view interactively rather than relying only on a text prompt.

Who it is for

  • People who want to create alternate camera views from an uploaded image.
  • Developers looking for a Python, Gradio and Three.js example of viewpoint-controlled image editing.

Documented uses

  • Set a desired camera direction and distance before generating an alternate image view.
  • Create several selected viewpoints in one batch.
  • Browse and download generated results from the current session.
  • Repeat a generation using a fixed seed.
Before you choose

Requirements & limits

The README describes the API route as requiring an Alibaba Cloud account and API key, with per-call billing. Its local-model route avoids API-call fees but requires a large model download and specified GPU and disk capacity.

  • The camera exposes eight azimuths and four elevations; directly overhead and underneath angles are not supported.
  • The README identifies Qwen-Image-Edit-Plus as the default API backend and says other models may not reproduce precise angles as accurately.
  • The project lists Python 3.10+, Gradio 4.x+ and Three.js r128 in its README badges.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Adapt the inference function for local Qwen models

Replace the API inference function with the README’s example local Qwen pipeline configuration.

Before you start

  • NVIDIA GPU with at least 16 GB VRAM; the README recommends RTX 3090 or above
  • At least 60 GB of free disk space
  • A roughly 56 GB model download
  • The listed Python dependencies, including diffusers, transformers, accelerate, peft, safetensors, sentencepiece and torchvision

Steps

  1. Install the repository requirements and the additional dependencies listed in the README.

  2. Replace infer_camera_edit in core.py with the documented local inference setup, which loads Qwen/Qwen-Image-Edit-2511 and the listed Lightning and Multiple-Angles LoRA weights.

  3. Start the interface with python app_zh.py or python app_en.py.

Expected outcome

The README’s example configures the application to use local Qwen image-editing inference.

Guide limitations

  • The README specifies substantial model storage and GPU memory requirements.
  • The example is a backend adaptation, not an API-free setup requiring no model download.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
ShuaixinHuang/image-multiple-angles-3d-camera · Official homepage title · retained in its source language
How we review sourcesSuggest a correction