zeromodels/xception41p_ra3_in1k

See our collection for all versions of Xception.

Run Xception with Keras 3: JAX, PyTorch, or TensorFlow

zeromodels/xception41p_ra3_in1k

Paper: Xception: Deep Learning with Depthwise Separable Convolutions (arXiv:1610.02357) · HF Papers

Xception interprets Inception modules as depthwise separable convolutions. Classifier or entry/middle/exit-flow backbone.

For more details on the model, please go to the upstream model card.

Pure-Keras 3 conversion of timm/xception41p.ra3_in1k for zeromodels. One implementation runs unmodified on TensorFlow / Torch / JAX.

This is an image-classification / backbone checkpoint (XceptionImageClassify / XceptionModel).

✨ Quick start

import os

os.environ["KERAS_BACKEND"] = "torch"  # or "jax" / "tensorflow"

from PIL import Image
from zeromodels.models.xception import XceptionImageClassify, XceptionModel, XceptionImageProcessor

model = XceptionImageClassify.from_weights("zeromodels/xception41p_ra3_in1k")
processor = XceptionImageProcessor.from_weights("zeromodels/xception41p_ra3_in1k")

image = Image.open("your_image.jpg").convert("RGB")
pixels = processor(image)  # resize + normalize (normalization lives in the processor)
logits = model(pixels, training=False)
print(logits.shape)  # (1, num_classes)

# Feature extraction: the backbone without the classifier head
backbone = XceptionModel.from_weights("zeromodels/xception41p_ra3_in1k", as_backbone=True)
features = backbone(pixels, training=False)

Load any Xception variant the same way with from_weights("zeromodels/<variant>"):

Variant	Hub
xception41_tf_in1k	zeromodels/xception41_tf_in1k
xception41p_ra3_in1k	zeromodels/xception41p_ra3_in1k
xception65_ra3_in1k	zeromodels/xception65_ra3_in1k
xception65_tf_in1k	zeromodels/xception65_tf_in1k
xception65p_ra3_in1k	zeromodels/xception65p_ra3_in1k
xception71_tf_in1k	zeromodels/xception71_tf_in1k

Tips

• Set KERAS_BACKEND before importing Keras / zeromodels.
• XceptionImageClassify returns class logits; XceptionModel returns features (as_backbone=True for multi-scale stages).
• See docs and Loading Weights.
• Upstream / timm checkpoints: XceptionImageClassify.from_weights("hf:timm/xception41p.ra3_in1k").

Special Thanks

A huge thank you to the Xception authors and the timm / Hub communities for creating and releasing these models.

License: see YAML license (usually matches the upstream checkpoint).