(target-models)=
# Models

Once you have labelled some trials by hand, a model can label the rest — and
you then curate what it predicted. Which model depends on **the shape of the
label** and on **what data exists when the model runs**.

```{mermaid}
flowchart TD
    start([Hand-labelled trials]) --> shape{What kind of label?}

    shape -->|"State event<br/>(an interval: onset + offset)"| segment[Action segmentation<br/><code>eto.segment</code><br/><i>DLC2Action models + added architectures</i>]
    shape -->|"Point event<br/>(one moment per trial)"| inference{What is available<br/>when the model runs?}

    inference -->|"Pose / time-series features,<br/>and you want to stay in the GUI"| lgbm[LightGBM model<br/><i>Model ▸ LightGBM: Train… / Predict…</i>]
    inference -->|Video only| spot[Event spotting: E2E-Spot<br/><code>eto.spot</code>]
    inference -->|Video + pose features| spotfeat[E2E-Spot + features]
    inference -->|"Video now,<br/>pose only for training"| distil[Pose teacher,<br/>distilled into E2E-Spot]

    segment --> curate
    lgbm --> curate
    spot --> curate
    spotfeat --> curate
    distil --> curate
    curate([Curate the predictions in the GUI]) --> conf[Read and threshold confidence]

    click segment "segment/index.html"
    click lgbm "onset_model.html"
    click spot "spot/index.html"
    click spotfeat "spot/multimodal.html"
    click distil "spot/multimodal.html"
    click curate "curation.html"
    click conf "confidence.html"
```

::::

## Where the models come from

- **LightGBM** {cite:p}`ke2017lightgbm`.
- **Action segmentation** — `eto.segment` vendors the models and loss of DLC2Action {cite:p}`kozlova2025dlc2action`, adapted there to pose/kinematic input:
  - DLC2Action's own variants: `mstcn` (MS-TCN3, which feeds the last two layers of the first stage into the second; from MS-TCN++ {cite:p}`li2020mstcnpp`) and `c2f_transformer` (C2F-TCN {cite:p}`singhania2021c2ftcn` with attention in place of convolution).
  - Original architectures, as adapted in DLC2Action: `asformer` {cite:p}`yi2021asformer`, `c2f_tcn` {cite:p}`singhania2021c2ftcn`, `edtcn` {cite:p}`lea2017edtcn`, `motionbert` {cite:p}`zhu2023motionbert`, and `mlp`, a per-frame baseline.
  - Added in EthoGraph: `rnn` (a bidirectional GRU/LSTM baseline), `specscalpel` {cite:p}`ji2026specscalpel` and `lady` {cite:p}`ji2026lady`.
- **Precise event spotting** — E2E-Spot {cite:p}`hong2022e2espot`.

```{toctree}
:maxdepth: 1
:hidden:

Action segmentation (state events) <segment/index>
LightGBM (point events from features) <onset_model>
PES (point events from pixels) <spot/index>
confidence
curation
```