How lsxi7/minima works
MINIMA combines an image-matching evaluation/demo runtime with a data engine that synthesizes six image modalities. The matching entry point loads a selected matcher, evaluates matches, and can render visualizations; the data-engine entry point dispatches modality transfers. Infrared transfer has a larger Scepter diffusion/inference implementation. README and source-index references establish the high-level capabilities and some links, but several modality and model internals were not sampled, so those connections are omitted.
Main entry points: modality_engine.py Modality Dispatcher · diffusion_inference.py Diffusion Inference · demo.py Matching Demo · stylebooth_inference.py StyleBooth Inference
Components 30
Image Matching
- Matching Demo
demo.py - Matcher Loader
load_model.py - Input Adapters
src/utils/data_io.py - Match Metrics
src/utils/metrics.py - Match Visualization
src/utils/plotting.py - Matcher Configs
src/config/default.py
Modality Data Engine
- Modality Dispatcher
data_engine/modality_engine.py - Depth Transfer
data_engine/tools/depth/depth_transfer.py - Event Transfer
data_engine/tools/event/event_transfer.py - Infrared Transfer
data_engine/tools/infrared/infrared_transfer.py - Normal Transfer
data_engine/tools/normal/normal_transfer.py - Paint Transfer
data_engine/tools/paint/paint_transfer.py - Sketch Transfer
data_engine/tools/sketch/sketch_transfer.py - DSINE Normal Models
data_engine/tools/normal/dsine/models/dsine/v02_kappa.py - Paint Network
data_engine/tools/paint/paint_transformer/network.py
Infrared Generation
- StyleBooth Inference
data_engine/tools/infrared/scepter/modules/inference/stylebooth_inference.py - Diffusion Inference
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - Control Inference
data_engine/tools/infrared/scepter/modules/inference/control_inference.py - Tuner Inference
data_engine/tools/infrared/scepter/modules/inference/tuner_inference.py
Scepter Internals
- Diffusion Model
data_engine/tools/infrared/scepter/modules/model/network/diffusion/diffusion.py - Diffusion Schedules
data_engine/tools/infrared/scepter/modules/model/network/diffusion/schedules.py - Diffusion Solvers
data_engine/tools/infrared/scepter/modules/model/network/diffusion/solvers.py - Model Registry
data_engine/tools/infrared/scepter/modules/model/registry.py - Configuration
data_engine/tools/infrared/scepter/modules/utils/config.py - Distributed Utilities
data_engine/tools/infrared/scepter/modules/utils/distribute.py - File System
data_engine/tools/infrared/scepter/modules/utils/file_system.py - Local File Client
data_engine/tools/infrared/scepter/modules/utils/file_clients/local_fs.py - Model Utilities
data_engine/tools/infrared/scepter/modules/utils/model.py - Model Probes
data_engine/tools/infrared/scepter/modules/utils/probe.py
Other components
- Engineer / User
Connections 30
- Engineer / User → Matching Demo: runs
README.md - Matching Demo → Matcher Loader: loads matcher
demo.py - Matching Demo → Match Metrics: evaluates matches
demo.py - Matching Demo → Match Visualization: renders results
demo.py - Matcher Loader → Input Adapters: wraps inputs
load_model.py - Matcher Loader → Matcher Configs: loads configuration
load_model.py - Engineer / User → Modality Dispatcher: chooses modality
data_engine/modality_engine.py - Modality Dispatcher → Depth Transfer: dispatches depth
data_engine/modality_engine.py - Modality Dispatcher → Event Transfer: dispatches event
data_engine/modality_engine.py - Modality Dispatcher → Infrared Transfer: dispatches infrared
data_engine/modality_engine.py - Modality Dispatcher → Normal Transfer: dispatches normal
data_engine/modality_engine.py - Modality Dispatcher → Paint Transfer: dispatches paint
data_engine/modality_engine.py - Modality Dispatcher → Sketch Transfer: dispatches sketch
data_engine/modality_engine.py - Infrared Transfer → StyleBooth Inference: runs inference
data_engine/tools/infrared/infrared_transfer.py - StyleBooth Inference → Diffusion Inference: delegates inference
data_engine/tools/infrared/scepter/modules/inference/stylebooth_inference.py - StyleBooth Inference → Control Inference: registers controls
data_engine/tools/infrared/scepter/modules/inference/stylebooth_inference.py - StyleBooth Inference → Tuner Inference: registers tuners
data_engine/tools/infrared/scepter/modules/inference/stylebooth_inference.py - Diffusion Inference → Diffusion Model: uses model
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - Diffusion Model → Diffusion Schedules: loads schedules
data_engine/tools/infrared/scepter/modules/model/network/diffusion/diffusion.py - Diffusion Model → Diffusion Solvers: uses solvers
data_engine/tools/infrared/scepter/modules/model/network/diffusion/diffusion.py - Diffusion Inference → Model Registry: builds models
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - Control Inference → Model Registry: builds controls
data_engine/tools/infrared/scepter/modules/inference/control_inference.py - Diffusion Inference → Configuration: reads configuration
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - Diffusion Inference → Distributed Utilities: uses distribution
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - Diffusion Inference → File System: loads files
data_engine/tools/infrared/scepter/modules/inference/diffusion_inference.py - File System → Local File Client: delegates storage
data_engine/tools/infrared/scepter/modules/utils/file_system.py - Configuration → Model Utilities: uses helpers
data_engine/tools/infrared/scepter/modules/utils/config.py - Distributed Utilities → Model Utilities: uses helpers
data_engine/tools/infrared/scepter/modules/utils/distribute.py - Normal Transfer → DSINE Normal Models: uses DSINE
data_engine/tools/normal/normal_transfer.py - Paint Transfer → Paint Network: uses network
data_engine/tools/paint/paint_transfer.py