-=HACKMANS=-

More AI VFX experiments with Wan2.1 VACE. Art directable explosions on existing footage plate. Being able to shape the effect and get all the free secondary motion of the environment interactions is phenomenal

Some AI VFX experiments to learn Wan2.1 VACE inpainting. We really are spoiled with the level of creative control possible from open source AI video

Tested #MultiTalk AI video (wan2.1) generation driven by audio, using a still image of Kurt and audio clip from Unplugged. Uncanny, but pretty impressed with the emotion and performance. 6 * 81 frame chunks, 47GB VRAM, 17min per chunk on RTX Pro 6000

4o gave up on this request, interesting progress image

Fantastic paper, SDXL ControlNet trained with Camera Intrinsic / Extrinsic parameters from 360 photos for controllable view and distortion in image+video generations. Pray for code + model release! graphics.unizar.es/projects/…

Creating some 'Talk Cycles' for my game @talesofsyn using #FluxSchnell txt2img from @bfl_ml -> #Gen3Turbo img2vid from @runwayml -> video mocap in @Cascadeur_soft #madewithunity #gamedev #indiedev #unity3d #animation

CFG testing on my Flux LoRA github.com/IdiotSandwichTheT…

First Flux.1-Dev @bfl_ml LoRA training test, not a style as such but more of a vibe leveraging prior knowledge but pushing more towards painterly cyberpunk #AI #Flux #LoRA

Got this beautiful glitch out of @runwayml #Gen3 from a regular prompt. It's 10x the file size and data rate of the regular video generations (SIEZURE WARNING flashing)