Loading video...
Video Failed to Load
Tried a mesh decal alternative for rendering damage effects. * Does't require writing data to RTs or doing per pixel checks for each hit location. * Can be rendered as overlay mesh or written into DBuffer for material blending :) #UE5 #Niagara
48,232 views • 1 year ago •via X (Twitter)
12 Comments
Skinning the extra vertices a second time, no?
Niagara can read from skin cache, so no skinning required :)
Microsoft and other top Nvidia AI chip customers have cut some orders of Blackwell AI chip server racks due to overheating.
Tutorial on this ?
Ooh, any resources or tips on your approach?
It's based on my NiagaraProceeduralMesh plugin. You can loop over triangles in mesh and add to procedural mesh if it passes distance/overlap check with decal positions. You can also write additional data in vertex color and texcoords to sample texture/opacity mask.
🤘🤘🤘
Looks good! What about performance, how does it work with many characters?
Each character would probably require a unique Niagara system. Not collected any stats on the setup, but maybe it would be quicker for a lot of localized hit locations (not covering a large area as that would increase procedural mesh gen cost)
Is there a way to sample the underlying material? Such as using the height data to help "fill" in the damage.
Looks cool! But for most projectiles it's using raytraces, why not get the hit location, surface normal and then spawn a deferred decal at that location parented to a bone socket? I'm trying it in my game but maybe it's really unoptimized 😄
Not sure if decals would perform too bad (increased draw call + CPU cost for updating decals). But I think deferred decal could look 'floaty' since the collision trace is done against physics collider (not actual surface) + wont follow mesh deformation?
![[NeurIPS '24] DreamMesh4D: Video-to-4D Generation with Sparse-Controlled Gaussian-Mesh Hybrid Representation Abstract (excerpt) We introduce DreamMesh4D, a novel framework that combines mesh representation with sparse-controlled deformation technique to generate high-quality 4D object from a monocular video. To overcome the limitation of classical texture representation, we bind Gaussian splats to the surface of the triangular mesh for differentiable optimization of both the texture and mesh vertices. In particular, DreamMesh4D begins with a coarse mesh provided by a single image based 3D generation method. Sparse points are then uniformly sampled across the surface of the mesh, and are used to build a deformation graph to drive the motion of the 3D object for the sake of computational efficiency and providing additional constraint. For each step, transformations of sparse control points are predicted using a deformation network, and the mesh vertices as well as the bound surface Gaussians are deformed via a geometric skinning algorithm. The skinning algorithm is a hybrid approach combining LBS (linear blending skinning) and DQS (dual-quaternion skinning), mitigating drawbacks associated with both approaches. The static surface Gaussians and mesh vertices as well as the dynamic deformation network are learned via reference view photometric loss, score distillation loss as well as other regularization losses in a two-stage manner. Extensive experiments demonstrate that our method outperforms prior video-to-4D generation methods in terms of rendering quality and spatial-temporal consistency.](https://image.24vids.com/tw-1843954158421840334/ext_tw_video_thumb/1843953495491174400/pu/img/Taktw8b95dHIcuGo.jpg)
