tomas-gajarsky/facetorch-fer-efficientnet-b0

EfficientNet-B0 facial-expression classifier

EmotiEffLib EfficientNet-B0 AffectNet checkpoint with seven visible-expression categories.

This repository contains immutable model artifacts used by Facetorch. Use the packaged Facetorch manifest to select a revision and artifact; do not treat mutable main or older unlisted files as a release contract.

Contract

Field	Value
Model ID	fer-efficientnet-b0
Architecture	TF EfficientNet-B0 with average pooling
Input	244 by 244 RGB face crop
Output	Seven logits ordered as Anger, Disgust, Fear, Happiness, Neutral, Sadness, Surprise.
Dynamic shapes	Batch dimension 1 through 64.
Weights license	Apache-2.0

Preprocessing: Resize to 244 by 244 and apply ImageNet mean [0.485, 0.456, 0.406] and standard deviation [0.229, 0.224, 0.225].

Release artifacts

File	Format	Runtime	Devices	SHA-256
model-torch2.6.pt2	pt2	>=2.6, <2.7	cpu, cuda	10937b9a9cc544cac440f938e7d706e9ef0321cdc9fc345d61f7b70e6efa27c9
model-torch2.11.pt2	pt2	>=2.11, <2.12	cpu, cuda	eddeb006828d128c2817bad74a6d5ee85775863c8bbff472e03e1bfb14a59dc0
model.pt	torchscript	>=2.6, <2.12	cpu	39d8046b1fe3eb06d5edb094307f0cd465a40b7f8886d4b477ca4bbaf7cbb62e

Facetorch v1 supports the Torch 2.6 and 2.11 cohort files listed in its manifest. The legacy TorchScript object is CPU-only and requires the explicit legacy opt-in. Files from unsupported cohorts are not part of the v1 release contract.

Loading the manifest-selected artifact

import torch
from huggingface_hub import hf_hub_download
from facetorch.artifacts import get_model_manifest

MODEL_ID = "fer-efficientnet-b0"
device = "cuda" if torch.cuda.is_available() else "cpu"
artifact = get_model_manifest().candidates(
    MODEL_ID,
    torch_version=torch.__version__,
    device=device,
    allow_legacy_models=False,
)[0]
path = hf_hub_download(
    repo_id=artifact.repo_id,
    revision=artifact.revision,
    filename=artifact.filename,
)
model = torch.export.load(path).module().to(device).eval()
example = torch.randn(1, 3, 244, 244, device=device)
with torch.inference_mode():
    output = model(example)

The random tensor above is only a loading smoke test. Use Facetorch's documented preprocessing for meaningful inference.

Provenance

Upstream	Immutable revision	Role	License
https://github.com/sb-ai-lab/EmotiEffLib	520a051c64cd191521e5934655314e769a319684	checkpoint publisher and architecture source	Apache-2.0

Upstream checkpoint	SHA-256	Source
models/affectnet_emotions/enet_b0_7.pt	772abad3fc90333ca3f5d454d20857900cad0a8988d1564d3ec6215673624010	publisher location

Mapping method: exact_tensor_equality_against_historical_author_repository_checkpoint.

Result: 360 of 360 tensors matched exactly.

The repository owner approved the mapping and redistribution record on 2026-08-23. Under the recorded policy, an author-published checkpoint in a permissively licensed repository with no separate checkpoint terms uses that repository license. MIT and Apache-2.0 have not been converted or treated as interchangeable. See LICENSE, THIRD_PARTY_NOTICES.md, and Facetorch's facetorch/models/governance.json.

Papers

• Classifying Emotions and Engagement in Online Learning Based on a Single Facial Expression Recognition Neural Network

Intended use

• Research classification of visible facial-expression categories.

Limitations and responsible use

• Expression labels do not establish emotion, intent, mental state, or truthfulness.
• Performance varies with culture, context, demographic representation, pose, and image quality.
• The checkpoint license does not grant rights to AffectNet or other training datasets.
• The artifact license does not itself license training datasets, input data, or a deployment's processing of personal data.
• Do not use model output as the sole basis for consequential decisions about a person.