Home / Open Source / Image and Video
All Image and Video items under Open Source on AIFIRST News, newest first. 9 items.
Media Grid is a free browser-based tool for splitting images into storyboard grids and creating character reference sheets, built with zero external dependencies. It targets creators working with AI-generated images who need visual organization tools for production pipelines.
GodEye is an open-source Android application that uses on-device AI to analyze live camera scenes, identify objects and situations, and deliver voice alerts in real time. The system targets surveillance and accessibility use cases where continuous scene description is needed without cloud round-trips. All processing runs locally on the device.
FitMirror lets users paste an Amazon, Flipkart, or Myntra product link and upload a photo to see the outfit rendered on themselves using AI image generation. The project is fully open-source and requires no account or paid API key.
A recent ComfyUI update modifies tokenizer handling in a way that changes how MiniMax H3 interprets prompt formatting. Users relying on established prompt templates for H3 may need to review and adjust their workflows after updating.
lylogummy released Anima-3.8B paired with Qwen-3.5 4B, a new image-video generation model available on Hugging Face. A dedicated ComfyUI custom node is provided for integration into standard generation workflows.
Fizgig v4.3.0 adds AMD Radeon support via ROCm, covering RDNA1 through RDNA4 GPUs, for LoRA training on Flux 2 Klein, Krea 2, and MiniMax H3. Previously NVIDIA-only, the free open-source tool now provides an AMD-native training path on Windows.
Python wrapper for ByteDance's Seedance 2.0 (delivered via muapi.ai). Supports text-to-video, image-to-video, video extension, character sheet generation, omni-reference (image+video+audio in one call). 2K output, 16:...
The VDN-H3 24GB ComfyUI node received a v1.1.0 update fixing adapter mapping issues that were causing degraded prompt adherence and memory problems. The developer corrected the root cause after identifying a discrepancy in the released adapter configuration.
A GitHub repository catalogues 7 node-based AI image generation tools with connection type, API details, pricing links, and live GitHub stats. The list targets ComfyUI-adjacent workflows and is maintained with current data.
AIFIRST 
