tue-mps/eomt

How tue-mps/eomt works

EoMT is a ViT-based image-segmentation research codebase: users configure training or validation, load dataset archives, and obtain instance, panoptic, or semantic predictions and metrics. The graph shows the documented CLI workflow, shared data and model layers, task-specific training branches, and the DINOv2/DINOv3 model choices. Several dataset and training relationships are visible in the source index, but not their full implementations; uncertain links are omitted rather than inferred.

Main entry points: datasets Dataset loaders · lightning_module.py Training and evaluation · main.py Train / validate CLI · coco_panoptic.py COCO panoptic

Components 24

CLI and configs

Train / validate CLI main.py
Task configs configs

Data pipeline

Dataset loaders datasets
Archive dataset datasets/dataset.py
Batch collation datasets/lightning_data_module.py
Image transforms datasets/transforms.py
COCO panoptic datasets/coco_panoptic.py
ADE20K panoptic datasets/ade20k_panoptic.py
COCO instance datasets/coco_instance.py
Semantic datasets datasets/ade20k_semantic.py
Cityscapes semantic datasets/cityscapes_semantic.py

Segmentation model

EoMT network models/eomt.py
ViT encoder models/vit.py
Mask upscaling models/scale_block.py

Training and tasks

Training and evaluation training/lightning_module.py
Instance task training/mask_classification_instance.py
Panoptic task training/mask_classification_panoptic.py
Semantic task training/mask_classification_semantic.py
Mask classification loss training/mask_classification_loss.py
Warmup polynomial schedule training/two_stage_warmup_poly_schedule.py

Model resources

Quick inference inference.ipynb
Pretrained model zoo model_zoo
DINOv2 / DINOv3 configs

Other components

Researcher

Connections 29

Researcher → Train / validate CLI: starts training or evaluation README.md
Train / validate CLI → Task configs: loads configuration README.md
Train / validate CLI → Dataset loaders: configures data module main.py
Train / validate CLI → Training and evaluation: configures model task README.md
Dataset loaders → COCO panoptic: provides dataset option datasets/coco_panoptic.py
Dataset loaders → ADE20K panoptic: provides dataset option datasets/ade20k_panoptic.py
Dataset loaders → COCO instance: provides dataset option datasets/coco_instance.py
Dataset loaders → Semantic datasets: provides dataset option datasets/ade20k_semantic.py
Dataset loaders → Cityscapes semantic: provides dataset option datasets/cityscapes_semantic.py
COCO panoptic → Batch collation: extends datasets/coco_panoptic.py
COCO panoptic → Archive dataset: constructs datasets/coco_panoptic.py
COCO panoptic → Image transforms: uses datasets/coco_panoptic.py
ADE20K panoptic → Batch collation: extends datasets/ade20k_panoptic.py
ADE20K panoptic → Archive dataset: constructs datasets/ade20k_panoptic.py
ADE20K panoptic → Image transforms: uses datasets/ade20k_panoptic.py
COCO instance → Archive dataset: uses datasets/coco_instance.py
COCO instance → Image transforms: uses datasets/coco_instance.py
EoMT network → Mask upscaling: uses models/eomt.py
EoMT network → ViT encoder: encodes image and queries
Training and evaluation → Warmup polynomial schedule: uses schedule training/lightning_module.py
Instance task → Training and evaluation: extends training/mask_classification_instance.py
Instance task → Mask classification loss: computes loss training/mask_classification_instance.py
Panoptic task → Training and evaluation: extends training/mask_classification_panoptic.py
Panoptic task → Mask classification loss: uses loss training/mask_classification_panoptic.py
Semantic task → Training and evaluation: extends training/mask_classification_semantic.py
Semantic task → Mask classification loss: uses loss training/mask_classification_semantic.py
Researcher → Quick inference: runs quick inference README.md
Researcher → Pretrained model zoo: selects pretrained weights README.md
Task configs → DINOv2 / DINOv3: selects DINO variant README.md

More Jupyter Notebook diagrams: salesforceairesearch/agentlite · langchain-ai/deep-agents-from-scratch · commaai/comma2k19 · johnsonhk88/ai-bank-statement-document-automation-by-llm-and-personal-finanical-analysis-prediction · nishitpatel01/fake_news_detection

Drawn by GitDiagram on September 30, 2026 from the repository's file tree, README and source files. Read as Markdown ·