Maskformer Swin Tiny Coco

🚀 MaskFormer

MaskFormer是一個在COCO全景分割數據集上訓練的模型（微型版本，採用Swin骨幹網絡）。它解決了實例分割、語義分割和全景分割等任務，為圖像分割領域帶來了新的解決方案。

🚀 快速開始

你可以使用以下代碼示例來使用這個模型進行語義分割：

from transformers import MaskFormerFeatureExtractor, MaskFormerForInstanceSegmentation
from PIL import Image
import requests

# load MaskFormer fine-tuned on COCO panoptic segmentation
feature_extractor = MaskFormerFeatureExtractor.from_pretrained("facebook/maskformer-swin-tiny-coco")
model = MaskFormerForInstanceSegmentation.from_pretrained("facebook/maskformer-swin-tiny-coco")

url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = feature_extractor(images=image, return_tensors="pt")

outputs = model(**inputs)
# model predicts class_queries_logits of shape `(batch_size, num_queries)`
# and masks_queries_logits of shape `(batch_size, num_queries, height, width)`
class_queries_logits = outputs.class_queries_logits
masks_queries_logits = outputs.masks_queries_logits

# you can pass them to feature_extractor for postprocessing
result = feature_extractor.post_process_panoptic_segmentation(outputs, target_sizes=[image.size[::-1]])[0]
# we refer to the demo notebooks for visualization (see "Resources" section in the MaskFormer docs)
predicted_panoptic_map = result["segmentation"]

更多代碼示例，請參考文檔。

✨ 主要特性

MaskFormer採用相同的範式來處理實例分割、語義分割和全景分割：通過預測一組掩碼和相應的標籤。因此，所有這3個任務都被視為實例分割任務。

📚 詳細文檔

預期用途和限制

你可以使用這個特定的檢查點進行語義分割。查看模型中心，以尋找其他針對你感興趣的任務進行微調的版本。

使用說明

這裡展示瞭如何使用這個模型：

from transformers import MaskFormerFeatureExtractor, MaskFormerForInstanceSegmentation
from PIL import Image
import requests

# 加載在COCO全景分割上微調的MaskFormer模型
feature_extractor = MaskFormerFeatureExtractor.from_pretrained("facebook/maskformer-swin-tiny-coco")
model = MaskFormerForInstanceSegmentation.from_pretrained("facebook/maskformer-swin-tiny-coco")

url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = feature_extractor(images=image, return_tensors="pt")

outputs = model(**inputs)
# 模型預測形狀為 `(batch_size, num_queries)` 的 class_queries_logits
# 以及形狀為 `(batch_size, num_queries, height, width)` 的 masks_queries_logits
class_queries_logits = outputs.class_queries_logits
masks_queries_logits = outputs.masks_queries_logits

# 你可以將它們傳遞給特徵提取器進行後處理
result = feature_extractor.post_process_panoptic_segmentation(outputs, target_sizes=[image.size[::-1]])[0]
# 我們參考演示筆記本進行可視化（請參閱MaskFormer文檔中的“資源”部分）
predicted_panoptic_map = result["segmentation"]

📄 許可證

許可證類型：other