Skill Claude CodeCodex
Removes predicted segmentation masks below a per-class minimum pixel area threshold to eliminate small false positive regions at inference time.
not rated 60
+1 4mo ago A 30 tokens
60Stars on the repository
200Mods indexed here, across every type
4mo agoLast push, which is what freshness is scored on
noneNo LICENSE: all rights reserved, so bodies are not copied
Skill Claude CodeCodex
Uses PyTorch AMP autocast and GradScaler for FP16 training, halving memory usage and speeding up training on modern GPUs.
not rated 60
+1 4mo ago A 33 tokens
Skill Claude CodeCodex
Combines mixup augmentation (linear interpolation of image pairs and their labels) with label smoothing in a single training pipeline for regularization.
not rated 60
+1 4mo ago A 35 tokens
Skill Claude CodeCodex
Route each DICOM series to a per-modality window-center / window-width pair (CT/CTA/MRA/MRI) before normalization, so the same model can ingest mixed modalities without one modality's intensity range washing out the others.
not rated 60
+1 4mo ago A 56 tokens
Skill Claude CodeCodex
Segment lung regions from CT using HU thresholding, connected-component labeling, and morphological opening.
not rated 60
+1 4mo ago A 27 tokens
Skill Claude CodeCodex
Merges overlapping same-class bounding boxes from multiple annotators into a consensus box using IoU-based matching and intersection.
not rated 60
+1 4mo ago A 34 tokens
Skill Claude CodeCodex
Combine EfficientNetB0..B6 into one Keras model with a shared image input and one sigmoid head per backbone, training all N models in a single fit() call on TPU.
not rated 60
+1 4mo ago A 46 tokens
Skill Claude CodeCodex
Generate count-regression training patches at a geometric pyramid of image scales (0.9^k) so one CNN handles within- and between-image object-size variation without explicit anchors.
not rated 60
+1 4mo ago A 45 tokens
Skill Claude CodeCodex
Stacks uniformly sampled slices from multiple MRI series (e.g., Sagittal T1, T2, Axial) into a single multi-channel tensor for one-pass inference.
not rated 60
+1 4mo ago A 43 tokens
Skill Claude CodeCodex
Stack multiple CT window settings (brain, subdural, bone) as separate RGB channels for CNN input.
not rated 60
+1 4mo ago A 29 tokens
Skill Claude CodeCodex
Composites multi-channel imagery (microscopy, satellite) into 3-channel RGB for pretrained CNN backbones.
not rated 60
+1 4mo ago A 33 tokens
Skill Claude CodeCodex
Computes per-class ROC-AUC then macro-averages for multi-label classification evaluation and model selection.
not rated 60
+1 4mo ago A 31 tokens
Skill Claude CodeCodex
For multi-label classification, compute the per-class binary correlation matrix restricted to multi-label rows and the conditional class counts given a rare anchor class — reveals label groupings the model can exploit (shared classifier heads, hierarchical loss weighting, post-hoc consistency rules).
not rated 60
+1 4mo ago A 64 tokens
Skill Claude CodeCodex
Oversample multi-label images by giving each image a duplication multiplier equal to the max per-class multiplier among its labels, so every rare class gets repetition without exploding common-class counts — the standard fix for long-tail multi-label distributions where SMOTE / per-row oversampling doesn't apply.
not rated 60
+1 4mo ago A 67 tokens
Skill Claude CodeCodex
Combine match predictions from image embeddings, text similarity, and perceptual hash via set union for maximum recall.
not rated 60
+1 4mo ago A 31 tokens
Skill Claude CodeCodex
UNet++ dense cross-depth skip connections that propagate deeper decoder features into all shallower decoder levels.
not rated 60
+1 4mo ago A 31 tokens
Skill Claude CodeCodex
Auto-detect whether a generated data series is numeric or categorical by measuring the fraction of digit characters in the concatenated values.
not rated 60
+1 4mo ago A 33 tokens
Skill Claude CodeCodex
Assign an unknown/novel class when all nearest-neighbor distances exceed a tuned cutoff threshold for open-set recognition.
not rated 60
+1 4mo ago A 29 tokens
Skill Claude CodeCodex
Uses Nelder-Mead optimization to find per-class decision thresholds that maximize Quadratic Weighted Kappa for regression-to-ordinal conversion.
not rated 60
+1 4mo ago A 35 tokens
Skill Claude CodeCodex
Encodes ordinal classes as cumulative binary labels (class N activates labels 0..N), enabling sigmoid + BCE training for ordinal regression.
not rated 60
+1 4mo ago A 37 tokens
Skill Claude CodeCodex
Uses packpaddedsequence to exclude padding tokens from cross-entropy loss in variable-length sequence generation.
not rated 60
+1 4mo ago A 28 tokens
Skill Claude CodeCodex
Compute Euclidean distance between entity pairs from tracking data and filter out pairs beyond a threshold to reduce inference candidates.
not rated 60
+1 4mo ago A 30 tokens
Skill Claude CodeCodex
Double left-join on tracking data to create pairwise features (positions, velocities, distance) for both entities in an interaction pair.
not rated 60
+1 4mo ago A 36 tokens
Skill Claude CodeCodex
Tile a large aerial image into fixed-size patches, accumulate per-class point-annotation counts into a grid tensor aligned with the tiles, and train a small CNN to regress per-class object counts per patch under MSE.
not rated 60
+1 4mo ago A 50 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: