How utat-ss/finch-science_syntheticdata works
FINCH-Science_SyntheticData develops conditional generators for hyperspectral spectra, synthesizes spectra under sampled or specified abundances, and evaluates whether the synthetic data improve unmixing and resemble real-data distributions. The principal workflow is training a generator, loading it for synthesis, then testing generated data with unmix/critic and nearest-neighbor metrics. The graph preserves the documented autoencoder, diffusion, and evaluation branches. Some subsystem wiring is supported only by the SOURCE INDEX or README; unsampled internals and relationships are not inferred.
Main entry points: loaders.py Synthesis model loaders · model.py CVAE model selection · unmix/script.py Unmix critic metric · master/script.py Evaluation orchestrator
Components 28
Generator training
- Diffusion data prep
defs/diffusion/data/data_preperation.py - Diffusion training
defs/diffusion/train.py - VAE data prep
defs/vae/data/data_preparation.py - VAE training
defs/vae/train.py - CVAE training entry
defs/hsi-cvae/main.py
Generator models
- CVAE model selection
defs/hsi-cvae/model.py - Dual-path CVAE
defs/hsi-cvae/models/dual_path_transformer/cvae.py - Conformer CVAE
defs/hsi-cvae/models/conformer/cvae.py - CVAE losses
defs/hsi-cvae/models/losses.py - VAE autoencoder
defs/vae/ae.py - TCVAE model
defs/tcvae/tcvae_model.py - TCVAE backbone
defs/tcvae/tcvae_backbone.py - Gaussian diffusion
defs/diffusion/diffusion.py - Conditional epsilon U-Net
defs/diffusion/epsilon/unet.py - Conformer components
defs/diffusion/epsilon/auxiliary/epsilon_addons.py - Noise schedules
defs/diffusion/noise/noise_scheduling.py
Data synthesis
- Synthesis entry
defs/synthesis/script.py - Synthesis model loaders
defs/synthesis/loaders.py - Abundance sampling
defs/synthesis/abundance_sampler.py - Spectrum samplers
defs/synthesis/synth_func.py - Lean synthesis
defs/synthesis/lean/lean_synth.py
Synthetic data evaluation
- Evaluation orchestrator
defs/testing/master/script.py - Nearest-neighbor metric
defs/testing/nearest_neighbor/script.py - Unmix critic metric
defs/testing/unmix/script.py - Critic data preparation
defs/testing/unmix/data/data.py - Critic training
defs/testing/unmix/critic_train/train.py - Critic model loaders
defs/testing/unmix/loaders/loaders.py
Other components
- Research user
Connections 29
- Research user → CVAE training entry: starts training
- Research user → Diffusion training: starts training
- Research user → VAE training: starts training
- Research user → Synthesis entry: starts synthesis
- Research user → Evaluation orchestrator: starts testing
- CVAE training entry → CVAE model selection: uses model
defs/hsi-cvae/main.py - CVAE model selection → Dual-path CVAE: selects architecture
defs/hsi-cvae/model.py - CVAE model selection → Conformer CVAE: selects architecture
defs/hsi-cvae/model.py - CVAE model selection → CVAE losses: uses losses
defs/hsi-cvae/model.py - TCVAE model → TCVAE backbone: uses backbone
defs/tcvae/tcvae_model.py - Gaussian diffusion → Noise schedules: uses schedule
defs/diffusion/diffusion.py - Conditional epsilon U-Net → Conformer components: uses components
defs/diffusion/epsilon/unet.py - Diffusion data prep → Diffusion training: prepares data
- VAE data prep → VAE training: prepares data
- Synthesis entry → Synthesis model loaders: loads models
- Synthesis model loaders → Gaussian diffusion: loads generator
defs/synthesis/loaders.py - Synthesis model loaders → VAE autoencoder: loads generator
defs/synthesis/loaders.py - Synthesis model loaders → TCVAE model: loads generator
defs/synthesis/loaders.py - Synthesis model loaders → Abundance sampling: uses sampler
defs/synthesis/loaders.py - Synthesis model loaders → Spectrum samplers: uses samplers
defs/synthesis/loaders.py - Spectrum samplers → Abundance sampling: uses sampling
defs/synthesis/synth_func.py - Lean synthesis → Spectrum samplers: uses samplers
defs/synthesis/lean/lean_synth.py - Evaluation orchestrator → Nearest-neighbor metric: dispatches metric
defs/testing/master/script.py - Evaluation orchestrator → Unmix critic metric: dispatches metric
defs/testing/master/script.py - Unmix critic metric → Critic data preparation: prepares data
defs/testing/unmix/script.py - Unmix critic metric → Critic model loaders: loads critic
defs/testing/unmix/script.py - Unmix critic metric → Critic training: trains critic
defs/testing/unmix/script.py - Critic training → Critic data preparation: uses data
defs/testing/unmix/critic_train/train.py - Critic model loaders → Critic training: configures training