nodes/SOLRICKS-comfyui-solricks
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

feat: smart neural pass

Browse Source
This commit is contained in:
MSXYZ-GenAI
2026-04-23 01:01:58 +03:00
parent 781d8f86f1
commit d89f4e0409
1 changed files with 25 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files
VideoTAADLAA.py
+25 -19
View File
@@ -1,6 +1,6 @@
# Created by MSXYZ (AI-assisted)
# Temporal (TAA) + DLAA Anti-Aliasing Inference
# v0.1.1 - Auto Tile Size
# v0.1.1 - Smart Neural Pass
import torch
import torch.nn as nn
@@ -127,15 +127,10 @@ class VideoTAADLAA:
return {
"required": {
"images" : ("IMAGE",),
"taa_strength" : ("FLOAT", {"default": 0.60, "min": 0, "max": 1, "step": 0.05}),
"taa_alpha" : ("FLOAT", {"default": 0.10, "min": 0, "max": 0.9, "step": 0.01}),
"motion_sensitivity": ("FLOAT", {"default": 0.08, "min": 0, "max": 0.3, "step": 0.01}),
"jitter_scale" : ("FLOAT", {"default": 0.20, "min": 0, "max": 1, "step": 0.01}),
"dlaa_strength" : ("FLOAT", {"default": 0.80, "min": 0, "max": 1.5, "step": 0.05}),
"sharpen_strength" : ("FLOAT", {"default": 0.30, "min": 0, "max": 2, "step": 0.05}),
"edge_threshold" : ("FLOAT", {"default": 0.15, "min": 0.05, "max": 0.5, "step": 0.01}),
"blur_radius" : ("INT", {"default": 1, "min": 0, "max": 3, "step": 1}),
"reset_history" : ("BOOLEAN", {"default": False}),
"taa_strength" : ("FLOAT", {"default": 0.50, "min": 0, "max": 1, "step": 0.05}),
"dlaa_strength" : ("FLOAT", {"default": 0.60, "min": 0, "max": 1.5, "step": 0.05}),
"sharpen_strength" : ("FLOAT", {"default": 0.20, "min": 0, "max": 2, "step": 0.05}),
"motion_sensitivity": ("FLOAT", {"default": 0.10, "min": 0, "max": 0.3, "step": 0.01}),
}
}
@@ -166,11 +161,8 @@ class VideoTAADLAA:
return self.net_cache[key]
def _tiled_forward(self, net, x: torch.Tensor, tile_size: int = 512, overlap: int = 32) -> torch.Tensor:
"""
Splits x into overlapping tiles, runs net on each tile,
blends with linear weight ramps to avoid seams.
VRAM usage stays constant regardless of input resolution.
"""
# Splits x into overlapping tiles to reduce VRAM usage
B, C, H, W = x.shape
# Small image
@@ -242,13 +234,27 @@ class VideoTAADLAA:
blurred = F.avg_pool2d(F.pad(x, [blur]*4, mode="reflect"), blur*2+1, stride=1)
return x*(1.0 - mask) + blurred*mask
def execute(self, images, taa_strength, taa_alpha, motion_sensitivity,
jitter_scale, dlaa_strength, sharpen_strength,
edge_threshold, blur_radius, reset_history=False):
def execute(self, images, taa_strength, motion_sensitivity, dlaa_strength, sharpen_strength):
device = mm.get_torch_device() if mm else \
torch.device("cuda" if torch.cuda.is_available() else "cpu")
# Fix
taa_alpha = 0.10
jitter_scale = 0.20
edge_threshold = 0.15
# IMAGE or VIDEO
B, H, W, C = images.shape
is_single_image = (B == 1)
if is_single_image:
blur_radius = 0
reset_history = True
else:
blur_radius = 1
reset_history = False
if reset_history:
self.taa.reset()
@@ -319,7 +325,7 @@ class VideoTAADLAA:
else:
raise e
# Orijinal görüntü ile DLAA sonucunu harmanla
# Combine the original image with the DLAA
rgb = torch.lerp(rgb, refined, dlaa_strength)
rgb = torch.clamp(rgb, 0.0, 1.0)