μ½˜ν…μΈ λ‘œ κ±΄λ„ˆλ›°κΈ°

Learn to Export to TFLite Edge TPU Format From YOLO11 Model

λͺ¨λ°”일 λ˜λŠ” μž„λ² λ””λ“œ μ‹œμŠ€ν…œκ³Ό 같이 μ—°μ‚° λŠ₯λ ₯이 μ œν•œλœ λ””λ°”μ΄μŠ€μ— 컴퓨터 λΉ„μ „ λͺ¨λΈμ„ λ°°ν¬ν•˜λŠ” 것은 κΉŒλ‹€λ‘œμšΈ 수 μžˆμŠ΅λ‹ˆλ‹€. 더 λΉ λ₯Έ μ„±λŠ₯에 μ΅œμ ν™”λœ λͺ¨λΈ ν˜•μ‹μ„ μ‚¬μš©ν•˜λ©΄ ν”„λ‘œμ„ΈμŠ€κ°€ κ°„μ†Œν™”λ©λ‹ˆλ‹€. TensorFlow Lite Edge TPU λ˜λŠ” TFLite Edge TPU λͺ¨λΈ ν˜•μ‹μ€ 신경망에 λΉ λ₯Έ μ„±λŠ₯을 μ œκ³΅ν•˜λ©΄μ„œ μ΅œμ†Œν•œμ˜ μ „λ ₯을 μ‚¬μš©ν•˜λ„λ‘ μ„€κ³„λ˜μ—ˆμŠ΅λ‹ˆλ‹€.

The export to TFLite Edge TPU format feature allows you to optimize your Ultralytics YOLO11 models for high-speed and low-power inferencing. In this guide, we'll walk you through converting your models to the TFLite Edge TPU format, making it easier for your models to perform well on various mobile and embedded devices.

TFLite Edge TPU 둜 내보내야 ν•˜λŠ” 이유 ?

Exporting models to TensorFlow Edge TPU makes machine learning tasks fast and efficient. This technology suits applications with limited power, computing resources, and connectivity. The Edge TPU is a hardware accelerator by Google. It speeds up TensorFlow Lite models on edge devices. The image below shows an example of the process involved.

TFLite Edge TPU

The Edge TPU works with quantized models. Quantization makes models smaller and faster without losing much accuracy. It is ideal for the limited resources of edge computing, allowing applications to respond quickly by reducing latency and allowing for quick data processing locally, without cloud dependency. Local processing also keeps user data private and secure since it's not sent to a remote server.

TFLite Edge의 μ£Όμš” κΈ°λŠ₯ TPU

λ‹€μŒμ€ TFLite Edge TPU λ₯Ό κ°œλ°œμžμ—κ²Œ ν›Œλ₯­ν•œ λͺ¨λΈ ν˜•μ‹ μ„ νƒμœΌλ‘œ λ§Œλ“œλŠ” μ£Όμš” κΈ°λŠ₯μž…λ‹ˆλ‹€:

  • 엣지 λ””λ°”μ΄μŠ€μ—μ„œ μ΅œμ ν™”λœ μ„±λŠ₯: TFLite Edge TPU λŠ” μ–‘μžν™”, λͺ¨λΈ μ΅œμ ν™”, ν•˜λ“œμ›¨μ–΄ 가속, 컴파일러 μ΅œμ ν™”λ₯Ό 톡해 고속 λ‰΄λŸ΄ λ„€νŠΈμ›Œν‚Ή μ„±λŠ₯을 λ‹¬μ„±ν•©λ‹ˆλ‹€. μ΅œμ†Œν•œμ˜ μ•„ν‚€ν…μ²˜λ‘œ 더 μž‘μ€ 크기와 λΉ„μš© νš¨μœ¨μ„±μ„ μ œκ³΅ν•©λ‹ˆλ‹€.

  • 높은 μ—°μ‚° μ²˜λ¦¬λŸ‰: TFLite Edge TPU λŠ” 특수 ν•˜λ“œμ›¨μ–΄ 가속과 효율적인 λŸ°νƒ€μž„ 싀행을 κ²°ν•©ν•˜μ—¬ 높은 μ»΄ν“¨νŒ… μ²˜λ¦¬λŸ‰μ„ λ‹¬μ„±ν•©λ‹ˆλ‹€. μ—„κ²©ν•œ μ„±λŠ₯ μš”κ±΄μ„ κ°–μΆ˜ λ¨Έμ‹  λŸ¬λ‹ λͺ¨λΈμ„ 엣지 λ””λ°”μ΄μŠ€μ— λ°°ν¬ν•˜λŠ” 데 μ ν•©ν•©λ‹ˆλ‹€.

  • Efficient Matrix Computations: The TensorFlow Edge TPU is optimized for matrix operations, which are crucial for neural network computations. This efficiency is key in machine learning models, particularly those requiring numerous and complex matrix multiplications and transformations.

TFLite Edgeλ₯Ό μ‚¬μš©ν•œ 배포 μ˜΅μ…˜ TPU

Before we jump into how to export YOLO11 models to the TFLite Edge TPU format, let's understand where TFLite Edge TPU models are usually used.

TFLite Edge( TPU )λŠ” λ‹€μŒκ³Ό 같은 λ¨Έμ‹  λŸ¬λ‹ λͺ¨λΈμ„ μœ„ν•œ λ‹€μ–‘ν•œ 배포 μ˜΅μ…˜μ„ μ œκ³΅ν•©λ‹ˆλ‹€:

  • μ˜¨λ””λ°”μ΄μŠ€ 배포: TensorFlow Edge TPU λͺ¨λΈμ€ λͺ¨λ°”일 및 μž„λ² λ””λ“œ λ””λ°”μ΄μŠ€μ— 직접 배포할 수 μžˆμŠ΅λ‹ˆλ‹€. μ˜¨λ””λ°”μ΄μŠ€ 배포λ₯Ό μ‚¬μš©ν•˜λ©΄ λͺ¨λΈμ„ ν•˜λ“œμ›¨μ–΄μ—μ„œ 직접 μ‹€ν–‰ν•  수 μžˆμœΌλ―€λ‘œ ν΄λΌμš°λ“œ 연결이 ν•„μš”ν•˜μ§€ μ•ŠμŠ΅λ‹ˆλ‹€.

  • ν΄λΌμš°λ“œλ₯Ό μ‚¬μš©ν•œ 엣지 μ»΄ν“¨νŒ… TensorFlow TPU: 엣지 λ””λ°”μ΄μŠ€μ˜ 처리 λŠ₯λ ₯이 μ œν•œμ μΈ μ‹œλ‚˜λ¦¬μ˜€μ—μ„œ TensorFlow 엣지 TPUλŠ” μΆ”λ‘  μž‘μ—…μ„ TPUκ°€ μž₯착된 ν΄λΌμš°λ“œ μ„œλ²„λ‘œ μ˜€ν”„λ‘œλ“œν•  수 μžˆμŠ΅λ‹ˆλ‹€.

  • Hybrid Deployment: A hybrid approach combines on-device and cloud deployment and offers a versatile and scalable solution for deploying machine learning models. Advantages include on-device processing for quick responses and cloud computing for more complex computations.

Exporting YOLO11 Models to TFLite Edge TPU

You can expand model compatibility and deployment flexibility by converting YOLO11 models to TensorFlow Edge TPU.

μ„€μΉ˜

ν•„μš”ν•œ νŒ¨ν‚€μ§€λ₯Ό μ„€μΉ˜ν•˜λ €λ©΄ μ‹€ν–‰ν•©λ‹ˆλ‹€:

μ„€μΉ˜

# Install the required package for YOLO11
pip install ultralytics

For detailed instructions and best practices related to the installation process, check our Ultralytics Installation guide. While installing the required packages for YOLO11, if you encounter any difficulties, consult our Common Issues guide for solutions and tips.

μ‚¬μš©λ²•

Before diving into the usage instructions, it's important to note that while all Ultralytics YOLO11 models are available for exporting, you can ensure that the model you select supports export functionality here.

μ‚¬μš©λ²•

from ultralytics import YOLO

# Load the YOLO11 model
model = YOLO("yolo11n.pt")

# Export the model to TFLite Edge TPU format
model.export(format="edgetpu")  # creates 'yolo11n_full_integer_quant_edgetpu.tflite'

# Load the exported TFLite Edge TPU model
edgetpu_model = YOLO("yolo11n_full_integer_quant_edgetpu.tflite")

# Run inference
results = edgetpu_model("https://ultralytics.com/images/bus.jpg")
# Export a YOLO11n PyTorch model to TFLite Edge TPU format
yolo export model=yolo11n.pt format=edgetpu  # creates 'yolo11n_full_integer_quant_edgetpu.tflite'

# Run inference with the exported model
yolo predict model=yolo11n_full_integer_quant_edgetpu.tflite source='https://ultralytics.com/images/bus.jpg'

μ§€μ›λ˜λŠ” 내보내기 μ˜΅μ…˜μ— λŒ€ν•œ μžμ„Έν•œ λ‚΄μš©μ€ 배포 μ˜΅μ…˜μ— λŒ€ν•œUltralytics λ¬Έμ„œ νŽ˜μ΄μ§€λ₯Ό μ°Έμ‘°ν•˜μ„Έμš”.

Deploying Exported YOLO11 TFLite Edge TPU Models

After successfully exporting your Ultralytics YOLO11 models to TFLite Edge TPU format, you can now deploy them. The primary and recommended first step for running a TFLite Edge TPU model is to use the YOLO("model_edgetpu.tflite") method, as outlined in the previous usage code snippet.

ν•˜μ§€λ§Œ TFLite Edge TPU λͺ¨λΈ 배포에 λŒ€ν•œ μžμ„Έν•œ 지침은 λ‹€μŒ λ¦¬μ†ŒμŠ€λ₯Ό μ°Έμ‘°ν•˜μ„Έμš”:

μš”μ•½

In this guide, we've learned how to export Ultralytics YOLO11 models to TFLite Edge TPU format. By following the steps mentioned above, you can increase the speed and power of your computer vision applications.

μ‚¬μš©λ²•μ— λŒ€ν•œ μžμ„Έν•œ λ‚΄μš©μ€ Edge 곡식 μ›Ήμ‚¬μ΄νŠΈ( TPU )μ—μ„œ ν™•μΈν•˜μ„Έμš”.

Also, for more information on other Ultralytics YOLO11 integrations, please visit our integration guide page. There, you'll discover valuable resources and insights.

자주 λ¬»λŠ” 질문

How do I export a YOLO11 model to TFLite Edge TPU format?

To export a YOLO11 model to TFLite Edge TPU format, you can follow these steps:

μ‚¬μš©λ²•

from ultralytics import YOLO

# Load the YOLO11 model
model = YOLO("yolo11n.pt")

# Export the model to TFLite Edge TPU format
model.export(format="edgetpu")  # creates 'yolo11n_full_integer_quant_edgetpu.tflite'

# Load the exported TFLite Edge TPU model
edgetpu_model = YOLO("yolo11n_full_integer_quant_edgetpu.tflite")

# Run inference
results = edgetpu_model("https://ultralytics.com/images/bus.jpg")
# Export a YOLO11n PyTorch model to TFLite Edge TPU format
yolo export model=yolo11n.pt format=edgetpu  # creates 'yolo11n_full_integer_quant_edgetpu.tflite'

# Run inference with the exported model
yolo predict model=yolo11n_full_integer_quant_edgetpu.tflite source='https://ultralytics.com/images/bus.jpg'

λ‹€λ₯Έ ν˜•μ‹μœΌλ‘œ λͺ¨λΈμ„ λ‚΄λ³΄λ‚΄λŠ” 방법에 λŒ€ν•œ μžμ„Έν•œ λ‚΄μš©μ€ 내보내기 κ°€μ΄λ“œλ₯Ό μ°Έμ‘°ν•˜μ„Έμš”.

What are the benefits of exporting YOLO11 models to TFLite Edge TPU?

Exporting YOLO11 models to TFLite Edge TPU offers several benefits:

  • μ΅œμ ν™”λœ μ„±λŠ₯: μ΅œμ†Œν•œμ˜ μ „λ ₯ μ†ŒλΉ„λ‘œ 고속 신경망 μ„±λŠ₯을 λ‹¬μ„±ν•˜μ„Έμš”.
  • 지연 μ‹œκ°„ 단좕: ν΄λΌμš°λ“œμ— μ˜μ‘΄ν•  ν•„μš” 없이 둜컬 데이터λ₯Ό λΉ λ₯΄κ²Œ μ²˜λ¦¬ν•©λ‹ˆλ‹€.
  • κ°•ν™”λœ κ°œμΈμ •λ³΄ 보호: 둜컬 처리λ₯Ό 톡해 μ‚¬μš©μž 데이터λ₯Ό λΉ„κ³΅κ°œλ‘œ μ•ˆμ „ν•˜κ²Œ λ³΄ν˜Έν•©λ‹ˆλ‹€.

This makes it ideal for applications in edge computing, where devices have limited power and computational resources. Learn more about why you should export.

λͺ¨λ°”일 및 μž„λ² λ””λ“œ λ””λ°”μ΄μŠ€μ— TFLite Edge TPU λͺ¨λΈμ„ 배포할 수 μžˆλ‚˜μš”?

예, TensorFlow Lite Edge TPU λͺ¨λΈμ€ λͺ¨λ°”일 및 μž„λ² λ””λ“œ λ””λ°”μ΄μŠ€μ— 직접 배포할 수 μžˆμŠ΅λ‹ˆλ‹€. 이 배포 방식을 μ‚¬μš©ν•˜λ©΄ λͺ¨λΈμ„ ν•˜λ“œμ›¨μ–΄μ—μ„œ 직접 μ‹€ν–‰ν•  수 μžˆμœΌλ―€λ‘œ 더 λΉ λ₯΄κ³  효율적인 좔둠이 κ°€λŠ₯ν•©λ‹ˆλ‹€. 톡합 μ˜ˆμ‹œλŠ” λΌμ¦ˆλ² λ¦¬νŒŒμ΄μ— Coral Edge 배포 κ°€μ΄λ“œ( TPU )λ₯Ό μ°Έμ‘°ν•˜μ„Έμš”.

TFLite Edge TPU λͺ¨λΈμ˜ 일반적인 μ‚¬μš© μ‚¬λ‘€λŠ” λ¬΄μ—‡μΈκ°€μš”?

TFLite Edge TPU λͺ¨λΈμ˜ 일반적인 μ‚¬μš© μ‚¬λ‘€λŠ” λ‹€μŒκ³Ό κ°™μŠ΅λ‹ˆλ‹€:

  • 슀마트 카메라: μ‹€μ‹œκ°„ 이미지 및 λΉ„λ””μ˜€ 뢄석 ν–₯상.
  • IoT λ””λ°”μ΄μŠ€: 슀마트 ν™ˆ 및 μ‚°μ—… μžλ™ν™” 지원.
  • ν—¬μŠ€μΌ€μ–΄: 의료 μ˜μƒ 및 진단 가속화.
  • μ†Œλ§€μ—…: 재고 관리 및 고객 행동 뢄석 κ°œμ„ .

μ΄λŸ¬ν•œ μ• ν”Œλ¦¬μΌ€μ΄μ…˜μ€ TFLite Edge TPU λͺ¨λΈμ˜ κ³ μ„±λŠ₯ 및 μ €μ „λ ₯ μ†ŒλΉ„μ˜ 이점을 λˆ„λ¦΄ 수 μžˆμŠ΅λ‹ˆλ‹€. μ‚¬μš© μ‹œλ‚˜λ¦¬μ˜€μ— λŒ€ν•΄ μžμ„Ένžˆ μ•Œμ•„λ³΄μ„Έμš”.

TFLite Edge TPU λͺ¨λΈμ„ λ‚΄λ³΄λ‚΄κ±°λ‚˜ λ°°ν¬ν•˜λŠ” λ™μ•ˆ 문제λ₯Ό ν•΄κ²°ν•˜λ €λ©΄ μ–΄λ–»κ²Œ ν•΄μ•Ό ν•˜λ‚˜μš”?

TFLite Edge TPU λͺ¨λΈμ„ λ‚΄λ³΄λ‚΄κ±°λ‚˜ λ°°ν¬ν•˜λŠ” λ™μ•ˆ λ¬Έμ œκ°€ λ°œμƒν•˜λŠ” 경우 일반적인 문제 κ°€μ΄λ“œλ₯Ό μ°Έμ‘°ν•˜μ—¬ 문제 ν•΄κ²° νŒμ„ ν™•μΈν•˜μ„Έμš”. 이 κ°€μ΄λ“œλŠ” μ›ν™œν•œ μž‘λ™μ„ 보μž₯ν•˜λŠ” 데 도움이 λ˜λŠ” 일반적인 λ¬Έμ œμ™€ ν•΄κ²° 방법을 λ‹€λ£Ήλ‹ˆλ‹€. μΆ”κ°€ 지원이 ν•„μš”ν•˜λ©΄ 도움말 μ„Όν„°λ₯Ό λ°©λ¬Έν•˜μ„Έμš”.


7κ°œμ›” μ „ 생성됨 ✏️ 11 일 μ „ μ—…λ°μ΄νŠΈλ¨

λŒ“κΈ€