Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. Weโ€™ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

๐ŸŒ [i18n-KO] Translated tasks/monocular_depth_estimation.mdx to Korean #23621

Merged
merged 1 commit into from
May 23, 2023
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions docs/source/ko/_toctree.yml
Original file line number Diff line number Diff line change
Expand Up @@ -62,8 +62,8 @@
title: ์ œ๋กœ์ƒท(zero-shot) ๊ฐ์ฒด ํƒ์ง€
- local: tasks/zero_shot_image_classification
title: ์ œ๋กœ์ƒท(zero-shot) ์ด๋ฏธ์ง€ ๋ถ„๋ฅ˜
- local: in_translation
title: (๋ฒˆ์—ญ์ค‘) Depth estimation
- local: tasks/monocular_depth_estimation
title: ๋‹จ์ผ ์˜์ƒ ๊ธฐ๋ฐ˜ ๊นŠ์ด ์ถ”์ •
title: (๋ฒˆ์—ญ์ค‘) ์ปดํ“จํ„ฐ ๋น„์ „
isExpanded: false
- sections:
Expand Down
145 changes: 145 additions & 0 deletions docs/source/ko/tasks/monocular_depth_estimation.mdx
Original file line number Diff line number Diff line change
@@ -0,0 +1,145 @@
<!--Copyright 2023 The HuggingFace Team. All rights reserved.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with
the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on
an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the
specific language governing permissions and limitations under the License.
-->

# ๋‹จ์ผ ์˜์ƒ ๊ธฐ๋ฐ˜ ๊นŠ์ด ์ถ”์ •[[depth-estimation-pipeline]]

๋‹จ์ผ ์˜์ƒ ๊ธฐ๋ฐ˜ ๊นŠ์ด ์ถ”์ •์€ ํ•œ ์žฅ๋ฉด์˜ ๋‹จ์ผ ์ด๋ฏธ์ง€์—์„œ ์žฅ๋ฉด์˜ ๊นŠ์ด ์ •๋ณด๋ฅผ ์˜ˆ์ธกํ•˜๋Š” ์ปดํ“จํ„ฐ ๋น„์ „ ์ž‘์—…์ž…๋‹ˆ๋‹ค.
์ฆ‰, ๋‹จ์ผ ์นด๋ฉ”๋ผ ์‹œ์ ์˜ ์žฅ๋ฉด์— ์žˆ๋Š” ๋ฌผ์ฒด์˜ ๊ฑฐ๋ฆฌ๋ฅผ ์˜ˆ์ธกํ•˜๋Š” ๊ณผ์ •์ž…๋‹ˆ๋‹ค.

๋‹จ์ผ ์˜์ƒ ๊ธฐ๋ฐ˜ ๊นŠ์ด ์ถ”์ •์€ 3D ์žฌ๊ตฌ์„ฑ, ์ฆ๊ฐ• ํ˜„์‹ค, ์ž์œจ ์ฃผํ–‰, ๋กœ๋ด‡ ๊ณตํ•™ ๋“ฑ ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์—์„œ ์‘์šฉ๋ฉ๋‹ˆ๋‹ค.
์กฐ๋ช… ์กฐ๊ฑด, ๊ฐ€๋ ค์ง, ํ…์Šค์ฒ˜์™€ ๊ฐ™์€ ์š”์†Œ์˜ ์˜ํ–ฅ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ๋Š” ์žฅ๋ฉด ๋‚ด ๋ฌผ์ฒด์™€ ํ•ด๋‹น ๊นŠ์ด ์ •๋ณด ๊ฐ„์˜ ๋ณต์žกํ•œ ๊ด€๊ณ„๋ฅผ ๋ชจ๋ธ์ด ์ดํ•ดํ•ด์•ผ ํ•˜๋ฏ€๋กœ ๊นŒ๋‹ค๋กœ์šด ์ž‘์—…์ž…๋‹ˆ๋‹ค.


<Tip>
์ด ํŠœํ† ๋ฆฌ์–ผ์—์„œ ๋‹ค๋ฃจ๋Š” ์ž‘์—…์€ ๋‹ค์Œ ๋ชจ๋ธ ์•„ํ‚คํ…์ฒ˜์—์„œ ์ง€์›๋ฉ๋‹ˆ๋‹ค:

<!--This tip is automatically generated by `make fix-copies`, do not fill manually!-->

[DPT](../model_doc/dpt), [GLPN](../model_doc/glpn)

<!--End of the generated tip-->

</Tip>

์ด๋ฒˆ ๊ฐ€์ด๋“œ์—์„œ ๋ฐฐ์šธ ๋‚ด์šฉ์€ ๋‹ค์Œ๊ณผ ๊ฐ™์Šต๋‹ˆ๋‹ค:

* ๊นŠ์ด ์ถ”์ • ํŒŒ์ดํ”„๋ผ์ธ ๋งŒ๋“ค๊ธฐ
* ์ง์ ‘ ๊นŠ์ด ์ถ”์ • ์ถ”๋ก ํ•˜๊ธฐ

์‹œ์ž‘ํ•˜๊ธฐ ์ „์—, ํ•„์š”ํ•œ ๋ชจ๋“  ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ๊ฐ€ ์„ค์น˜๋˜์–ด ์žˆ๋Š”์ง€ ํ™•์ธํ•˜์„ธ์š”:

```bash
pip install -q transformers
```

## ๊นŠ์ด ์ถ”์ • ํŒŒ์ดํ”„๋ผ์ธ[[depth-estimation-inference-by-hand]]

๊นŠ์ด ์ถ”์ •์„ ์ถ”๋ก ํ•˜๋Š” ๊ฐ€์žฅ ๊ฐ„๋‹จํ•œ ๋ฐฉ๋ฒ•์€ ํ•ด๋‹น ๊ธฐ๋Šฅ์„ ์ œ๊ณตํ•˜๋Š” [`pipeline`]์„ ์‚ฌ์šฉํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
[Hugging Face Hub ์ฒดํฌํฌ์ธํŠธ](https://huggingface.co/models?pipeline_tag=depth-estimation&sort=downloads)์—์„œ ํŒŒ์ดํ”„๋ผ์ธ์„ ์ดˆ๊ธฐํ™”ํ•ฉ๋‹ˆ๋‹ค:

```py
>>> from transformers import pipeline

>>> checkpoint = "vinvino02/glpn-nyu"
>>> depth_estimator = pipeline("depth-estimation", model=checkpoint)
```


๋‹ค์Œ์œผ๋กœ, ๋ถ„์„ํ•  ์ด๋ฏธ์ง€๋ฅผ ํ•œ ์žฅ ์„ ํƒํ•˜์„ธ์š”:

```py
>>> from PIL import Image
>>> import requests

>>> url = "https://unsplash.com/photos/HwBAsSbPBDU/download?ixid=MnwxMjA3fDB8MXxzZWFyY2h8MzR8fGNhciUyMGluJTIwdGhlJTIwc3RyZWV0fGVufDB8MHx8fDE2Nzg5MDEwODg&force=true&w=640"
>>> image = Image.open(requests.get(url, stream=True).raw)
>>> image
```

<div class="flex justify-center">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/depth-estimation-example.jpg" alt="Photo of a busy street"/>
</div>

์ด๋ฏธ์ง€๋ฅผ ํŒŒ์ดํ”„๋ผ์ธ์œผ๋กœ ์ „๋‹ฌํ•ฉ๋‹ˆ๋‹ค.

```py
>>> predictions = depth_estimator(image)
```

ํŒŒ์ดํ”„๋ผ์ธ์€ ๋‘ ๊ฐœ์˜ ํ•ญ๋ชฉ์„ ๊ฐ€์ง€๋Š” ๋”•์…”๋„ˆ๋ฆฌ๋ฅผ ๋ฐ˜ํ™˜ํ•ฉ๋‹ˆ๋‹ค.
์ฒซ ๋ฒˆ์งธ๋Š” `predicted_depth`๋กœ ๊ฐ ํ”ฝ์…€์˜ ๊นŠ์ด๋ฅผ ๋ฏธํ„ฐ๋กœ ํ‘œํ˜„ํ•œ ๊ฐ’์„ ๊ฐ€์ง€๋Š” ํ…์„œ์ž…๋‹ˆ๋‹ค.
๋‘ ๋ฒˆ์งธ๋Š” `depth`๋กœ ๊นŠ์ด ์ถ”์ • ๊ฒฐ๊ณผ๋ฅผ ์‹œ๊ฐํ™”ํ•˜๋Š” PIL ์ด๋ฏธ์ง€์ž…๋‹ˆ๋‹ค.

์ด์ œ ์‹œ๊ฐํ™”ํ•œ ๊ฒฐ๊ณผ๋ฅผ ์‚ดํŽด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค:

```py
>>> predictions["depth"]
```

<div class="flex justify-center">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/depth-visualization.png" alt="Depth estimation visualization"/>
</div>

## ์ง์ ‘ ๊นŠ์ด ์ถ”์ • ์ถ”๋ก ํ•˜๊ธฐ[[depth-estimation-inference-by-hand]]

์ด์ œ ๊นŠ์ด ์ถ”์ • ํŒŒ์ดํ”„๋ผ์ธ ์‚ฌ์šฉ๋ฒ•์„ ์‚ดํŽด๋ณด์•˜์œผ๋‹ˆ ๋™์ผํ•œ ๊ฒฐ๊ณผ๋ฅผ ๋ณต์ œํ•˜๋Š” ๋ฐฉ๋ฒ•์„ ์‚ดํŽด๋ณด๊ฒ ์Šต๋‹ˆ๋‹ค.
[Hugging Face Hub ์ฒดํฌํฌ์ธํŠธ](https://huggingface.co/models?pipeline_tag=depth-estimation&sort=downloads)์—์„œ ๋ชจ๋ธ๊ณผ ๊ด€๋ จ ํ”„๋กœ์„ธ์„œ๋ฅผ ๊ฐ€์ ธ์˜ค๋Š” ๊ฒƒ๋ถ€ํ„ฐ ์‹œ์ž‘ํ•ฉ๋‹ˆ๋‹ค.
์—ฌ๊ธฐ์„œ ์ด์ „์— ์‚ฌ์šฉํ•œ ์ฒดํฌํฌ์ธํŠธ์™€ ๋™์ผํ•œ ๊ฒƒ์„ ์‚ฌ์šฉํ•ฉ๋‹ˆ๋‹ค:

```py
>>> from transformers import AutoImageProcessor, AutoModelForDepthEstimation

>>> checkpoint = "vinvino02/glpn-nyu"

>>> image_processor = AutoImageProcessor.from_pretrained(checkpoint)
>>> model = AutoModelForDepthEstimation.from_pretrained(checkpoint)
```

ํ•„์š”ํ•œ ์ด๋ฏธ์ง€ ๋ณ€ํ™˜์„ ์ฒ˜๋ฆฌํ•˜๋Š” `image_processor`๋ฅผ ์‚ฌ์šฉํ•˜์—ฌ ๋ชจ๋ธ์— ๋Œ€ํ•œ ์ด๋ฏธ์ง€ ์ž…๋ ฅ์„ ์ค€๋น„ํ•ฉ๋‹ˆ๋‹ค.
`image_processor`๋Š” ํฌ๊ธฐ ์กฐ์ • ๋ฐ ์ •๊ทœํ™” ๋“ฑ ํ•„์š”ํ•œ ์ด๋ฏธ์ง€ ๋ณ€ํ™˜์„ ์ฒ˜๋ฆฌํ•ฉ๋‹ˆ๋‹ค:

```py
>>> pixel_values = image_processor(image, return_tensors="pt").pixel_values
```

์ค€๋น„ํ•œ ์ž…๋ ฅ์„ ๋ชจ๋ธ๋กœ ์ „๋‹ฌํ•ฉ๋‹ˆ๋‹ค:

```py
>>> import torch

>>> with torch.no_grad():
... outputs = model(pixel_values)
... predicted_depth = outputs.predicted_depth
```

๊ฒฐ๊ณผ๋ฅผ ์‹œ๊ฐํ™”ํ•ฉ๋‹ˆ๋‹ค:

```py
>>> import numpy as np

>>> # ์›๋ณธ ์‚ฌ์ด์ฆˆ๋กœ ๋ณต์›
>>> prediction = torch.nn.functional.interpolate(
... predicted_depth.unsqueeze(1),
... size=image.size[::-1],
... mode="bicubic",
... align_corners=False,
... ).squeeze()
>>> output = prediction.numpy()

>>> formatted = (output * 255 / np.max(output)).astype("uint8")
>>> depth = Image.fromarray(formatted)
>>> depth
```

<div class="flex justify-center">
<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/depth-visualization.png" alt="Depth estimation visualization"/>
</div>