refactor(tools): generalize infer.py to accept any registered model

[idchacon28]

Generalize tools/infer.py to accept any registered model

Why

tools/infer.py was hardcoded to the legacy multi-class HerdNet model:

• from animaloc.models import LossWrapper, HerdNet
• from animaloc.eval import HerdNetStitcher, HerdNetEvaluator
• assert num_classes == 7 (literal)
• hardcoded classes = {1:'class_1', ..., 6:'class_6'}

Result: single-class OWL models (OWLC, OWLT, OWLD_S/B/L/H) could not run inference through this script. The docs routed users to tools/test.py, which requires ground-truth annotations — useless for actual inference on new imagery.

What

Two commits do the work, plus tests and docs:

Commit	What
feat(registry): add animaloc/registry/families.py	New FAMILIES table mapping each registered model to its stitcher, evaluator, default model_kwargs, normalization stats, down_ratio, and multi_class flag. Plus a resolve_family(name, *, checkpoint_meta, overrides) helper. Lives under animaloc/registry/ so any future tool can import it without dragging eval deps into animaloc.models.
refactor(tools): generalize infer.py	Full rewrite. Replaces all HerdNet hardcoding with family-driven lookups. New CLI flags. Backwards-compatible default (--model HerdNet).
test(smoke): add tests/smoke_infer.sh	Bash one-shot that ensures the synthetic dataset + an OWL-C checkpoint exist, runs tools/infer.py --model OWLC, and verifies the CSV schema.
docs(infer): update training.md	Replaces the "for OWL inference use test.py" workaround with the real recipe.

New CLI surface

python tools/infer.py <images_dir> <model.pth>                # back-compat: HerdNet
python tools/infer.py <images_dir> <model.pth> --model OWLC
python tools/infer.py <images_dir> <model.pth> --model OWLD_L -device cpu \
    --output-dir /tmp/owld_l_results
python tools/infer.py <images_dir> <model.pth> --model OWLT \
    --model-kwarg down_ratio=4

Full flag set: --model, --model-kwarg KEY=VAL (repeatable), --stitcher, --evaluator, --num-classes, --mean, --std, --down-ratio, --lmds-kernel-size, --lmds-adapt-ts, --lmds-neg-ts, --output-dir. Plus the existing -size -over -device -pf -rot --skip-model-inference.

Resolution order (per setting)

1. FAMILIES[name] defaults
2. Checkpoint metadata (classes, mean, std saved by tools/train.py)
3. Explicit CLI override

model_kwargs is merged rather than replaced, so users override one kwarg without listing every default.

Behavior changes that matter

• Output directory is configurable (--output-dir) and the default folder name now mentions the model (20260605_OWLC_results, not the hardcoded _HerdNet_results). HerdNet default location is unchanged.
• assert num_classes == 7 is gone. For HerdNet without classes metadata, the layer-shape probe (model.cls_head.2.weight) is kept as a last-resort fallback. OWL families don't take num_classes.
• state_dict loading is now strict=False with explicit missing/unexpected key warnings (catches partial-load checkpoints; tells the user what happened).
• The .map(classes) + .dropna() chain is gone. Detection rows whose label is unmapped now keep the raw label (as string) in species and emit one warning listing unmapped labels.
• pretrained=False is set in every family's model_kwargs so the constructor never re-fetches DINOv3 or DLA-34 weights at inference time. The checkpoint's state_dict supersedes them.

Sanity checks added at startup

• One-shot dummy forward on torch.zeros(1, 3, size, size) to detect model/stitcher shape mismatches early. Unwraps LossWrapper.forward's (output, output_dict) tuple and ignores None entries in tuple outputs (e.g. OWLD_S returns (heatmap, None)).

Smoke validation (CPU, Python 3.11)

Test	Result
tools/infer.py val/ <OWLC_ckpt> --model OWLC -device cpu	✅ 1856 detections written to CSV
tools/infer.py val/ <OWLD_S_ckpt> --model OWLD_S -device cpu	✅ 2270 detections; full DINOv3 + animaloc end-to-end
tools/infer.py val/ <OWLC_ckpt> (default HerdNet against OWL ckpt)	✅ fails cleanly with "4 missing key(s)" warning + actionable LMDS error
./tests/smoke_infer.sh	✅ passes (290 detections, schema OK)
uv run mkdocs build --strict	✅ no warnings

Deferred to follow-up PRs

1. Clean up the duplicate class definitions in animaloc/eval/evaluators.py (the whole file is pasted twice; verified the first-half classes are operationally bound, but it's a mess).
2. Add model_name / stitcher_name / evaluator_name to tools/train.py's checkpoint metadata so --model could become fully auto-detected.
3. Factor out a pure inference path that doesn't use Evaluator with dummy ground truth.

Backwards compatibility

Old usage	After this PR
tools/infer.py images/ ckpt.pth	Still works. Defaults to --model HerdNet.
-size 512 -over 160 -device cuda (etc.)	Unchanged.
_HerdNet_results output folder for HerdNet	Unchanged (still uses _HerdNet_ since --model defaults to HerdNet).

The only intentional break is the assert num_classes == 7. Any legacy HerdNet checkpoint with a different num_classes will now actually load.

Branch

feat/generalize-infer — 4 commits, all conventional-commits style:

8347984 docs(infer): update training.md for the generalized tools/infer.py
85b66c0 test(smoke): add tests/smoke_infer.sh + extend tests/README
0ab5abc refactor(tools): generalize infer.py to support any registered model
8bb1d25 feat(registry): add animaloc/registry/families.py for deployment defaults

How to test locally

gh pr checkout <this-PR-number>
uv sync
uv run python tools/infer.py --help     # see the new CLI
./tests/smoke_infer.sh                  # end-to-end smoke test

[microsoft-github-policy-service]

@idchacon28 please read the following Contributor License Agreement(CLA). If you agree with the CLA, please reply with the following information.

@microsoft-github-policy-service agree [company="{your company}"]

Options:

• (default - no company specified) I have sole ownership of intellectual property rights to my Submissions and I am not making Submissions in the course of work for my employer.

@microsoft-github-policy-service agree

• (when company given) I am making Submissions in the course of work for my employer (or my employer has intellectual property rights in my Submissions by contract or applicable law). I have permission from my employer to make Submissions and enter into this Agreement on behalf of my employer. By signing below, the defined term “You” includes me and my employer.

@microsoft-github-policy-service agree company="Microsoft"

Contributor License Agreement

Contribution License Agreement

This Contribution License Agreement (“Agreement”) is agreed to by the party signing below (“You”),
and conveys certain license rights to Microsoft Corporation and its affiliates (“Microsoft”) for Your
contributions to Microsoft open source projects. This Agreement is effective as of the latest signature
date below.

1. Definitions.
   “Code” means the computer software code, whether in human-readable or machine-executable form,
   that is delivered by You to Microsoft under this Agreement.
   “Project” means any of the projects owned or managed by Microsoft and offered under a license
   approved by the Open Source Initiative (www.opensource.org).
   “Submit” is the act of uploading, submitting, transmitting, or distributing code or other content to any
   Project, including but not limited to communication on electronic mailing lists, source code control
   systems, and issue tracking systems that are managed by, or on behalf of, the Project for the purpose of
   discussing and improving that Project, but excluding communication that is conspicuously marked or
   otherwise designated in writing by You as “Not a Submission.”
   “Submission” means the Code and any other copyrightable material Submitted by You, including any
   associated comments and documentation.
2. Your Submission. You must agree to the terms of this Agreement before making a Submission to any
   Project. This Agreement covers any and all Submissions that You, now or in the future (except as
   described in Section 4 below), Submit to any Project.
3. Originality of Work. You represent that each of Your Submissions is entirely Your original work.
   Should You wish to Submit materials that are not Your original work, You may Submit them separately
   to the Project if You (a) retain all copyright and license information that was in the materials as You
   received them, (b) in the description accompanying Your Submission, include the phrase “Submission
   containing materials of a third party:” followed by the names of the third party and any licenses or other
   restrictions of which You are aware, and (c) follow any other instructions in the Project’s written
   guidelines concerning Submissions.
4. Your Employer. References to “employer” in this Agreement include Your employer or anyone else
   for whom You are acting in making Your Submission, e.g. as a contractor, vendor, or agent. If Your
   Submission is made in the course of Your work for an employer or Your employer has intellectual
   property rights in Your Submission by contract or applicable law, You must secure permission from Your
   employer to make the Submission before signing this Agreement. In that case, the term “You” in this
   Agreement will refer to You and the employer collectively. If You change employers in the future and
   desire to Submit additional Submissions for the new employer, then You agree to sign a new Agreement
   and secure permission from the new employer before Submitting those Submissions.
5. Licenses.

• Copyright License. You grant Microsoft, and those who receive the Submission directly or
  indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license in the
  Submission to reproduce, prepare derivative works of, publicly display, publicly perform, and distribute
  the Submission and such derivative works, and to sublicense any or all of the foregoing rights to third
  parties.
• Patent License. You grant Microsoft, and those who receive the Submission directly or
  indirectly from Microsoft, a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license under
  Your patent claims that are necessarily infringed by the Submission or the combination of the
  Submission with the Project to which it was Submitted to make, have made, use, offer to sell, sell and
  import or otherwise dispose of the Submission alone or with the Project.
• Other Rights Reserved. Each party reserves all rights not expressly granted in this Agreement.
  No additional licenses or rights whatsoever (including, without limitation, any implied licenses) are
  granted by implication, exhaustion, estoppel or otherwise.

6. Representations and Warranties. You represent that You are legally entitled to grant the above
   licenses. You represent that each of Your Submissions is entirely Your original work (except as You may
   have disclosed under Section 3). You represent that You have secured permission from Your employer to
   make the Submission in cases where Your Submission is made in the course of Your work for Your
   employer or Your employer has intellectual property rights in Your Submission by contract or applicable
   law. If You are signing this Agreement on behalf of Your employer, You represent and warrant that You
   have the necessary authority to bind the listed employer to the obligations contained in this Agreement.
   You are not expected to provide support for Your Submission, unless You choose to do so. UNLESS
   REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING, AND EXCEPT FOR THE WARRANTIES
   EXPRESSLY STATED IN SECTIONS 3, 4, AND 6, THE SUBMISSION PROVIDED UNDER THIS AGREEMENT IS
   PROVIDED WITHOUT WARRANTY OF ANY KIND, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTY OF
   NONINFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
7. Notice to Microsoft. You agree to notify Microsoft in writing of any facts or circumstances of which
   You later become aware that would make Your representations in this Agreement inaccurate in any
   respect.
8. Information about Submissions. You agree that contributions to Projects and information about
   contributions may be maintained indefinitely and disclosed publicly, including Your name and other
   information that You submit with Your Submission.
9. Governing Law/Jurisdiction. This Agreement is governed by the laws of the State of Washington, and
   the parties consent to exclusive jurisdiction and venue in the federal courts sitting in King County,
   Washington, unless no federal subject matter jurisdiction exists, in which case the parties consent to
   exclusive jurisdiction and venue in the Superior Court of King County, Washington. The parties waive all
   defenses of lack of personal jurisdiction and forum non-conveniens.
10. Entire Agreement/Assignment. This Agreement is the entire agreement between the parties, and
    supersedes any and all prior agreements, understandings or communications, written or oral, between
    the parties relating to the subject matter hereof. This Agreement may be assigned by Microsoft.