人脸检测
使用集合让一切井井有条
根据您的偏好保存内容并对其进行分类。
借助机器学习套件的人脸检测 API，您可以检测图片中的人脸、识别主要的面部特征，并获取检测到的人脸的轮廓。请注意，该 API 用于检测人脸，不识别人脸。
借助人脸检测，您可以获取执行任务（如美化自拍照和肖像，或根据用户的照片生成头像）所需的信息。由于机器学习套件可以实时执行人脸检测，因此您可以在视频聊天或响应玩家表情的游戏等应用中使用该功能。
iOS
Android
主要功能
- 识别和定位面部特征
获取检测到的每个人脸的眼睛、耳朵、脸颊、鼻子和嘴巴的坐标。
- 获取面部特征的轮廓
获取检测到的面部的轮廓及其眼睛、眉毛、嘴唇和鼻子。
- 识别面部表情
确定人物是在微笑还是闭着眼睛。
- 跨视频帧跟踪人脸
获取每个检测到的唯一身份人脸的标识符。
标识符在不同调用中保持一致，因此您可以对视频流中的特定人员执行图片处理。
- 实时处理视频帧：人脸检测在设备上执行，其速度足以在视频处理等实时应用中使用。
示例结果
示例 1
对于检测到的每个人脸：
| 第 1 个面孔（共 3 个） |
|---|
| 边界多边形 |
|
| 旋转角度 |
Y：-14.054030418395996，Z：-55.007488250732422 |
| 跟踪 ID |
2 |
| 面部特征点 |
| 左眼 |
（945.869323730469、211.867126464844） |
| 右眼 |
（971.579467773438、247.257247924805） |
| 嘴巴下部 |
（907.756591796875、259.714477539062） |
等等
|
| 特征概率 |
| 微笑 |
0.88979166746139526 |
| 左眼睁开 |
0.98635888937860727 |
| 右眼睁开 |
0.99258323386311531 |
|
示例 2（人脸轮廓检测）
启用人脸轮廓检测后，对于检测到的每个面部特征，您还会获得一系列点。这些点表示地图项的形状。下图说明了这些点如何映射到人脸。点击图片可放大：
| 面部特征轮廓 |
|---|
| 鼻梁 |
(505.149811, 221.201797)、(506.987122, 313.285919)
|
| 左眼 |
(404.642029, 232.854431), (408.527283, 231.366623), (413.565796, 229.427856), (421.378296, 226.967682), (432.598755, 225.434143), (442.953064, 226.089508), (453.899811, 228.594818), (461.516418, 232.650467), (465.069580, 235.600845), (462.170410, 236.316147), (456.233643, 236.891602), (446.363922, 237.966888), (435.698914, 238.149323), (424.320740, 237.235168), (416.037720, 236.012115), (409.983459, 234.870300)
|
| 上唇顶部 |
(421.662048, 354.520813), (428.103882, 349.694061), (440.847595, 348.048737), (456.549988, 346.295532), (480.526489, 346.089294), (503.375702, 349.470459), (525.624634, 347.352783), (547.371155, 349.091980), (560.082031, 351.693268), (570.226685, 354.210175), (575.305420, 359.257751)
|
| （等等） |
|
如未另行说明，那么本页面中的内容已根据知识共享署名 4.0 许可获得了许可，并且代码示例已根据 Apache 2.0 许可获得了许可。有关详情，请参阅 Google 开发者网站政策。Java 是 Oracle 和/或其关联公司的注册商标。
最后更新时间 (UTC)：2025-08-29。
[null,null,["最后更新时间 (UTC)：2025-08-29。"],[[["\u003cp\u003eML Kit's Face Detection API can detect faces and their features in images and videos, but it does not recognize individuals.\u003c/p\u003e\n"],["\u003cp\u003eIt can be used for tasks such as adding effects to photos, creating avatars, and building interactive applications that respond to facial expressions.\u003c/p\u003e\n"],["\u003cp\u003eThe API provides facial feature coordinates, contours, expression detection (like smiling), and face tracking across video frames.\u003c/p\u003e\n"],["\u003cp\u003eFace detection happens on the device and is fast enough for real-time applications.\u003c/p\u003e\n"]]],[],null,["With ML Kit's face detection API, you can detect faces in an image, identify\nkey facial features, and get the contours of detected faces. Note that the API\n*detects faces* , it does not *recognize people* .\n\nWith face detection, you can get the information you need to perform tasks like\nembellishing selfies and portraits, or generating avatars from a user's photo.\nBecause ML Kit can perform face detection in real time, you can use it in\napplications like video chat or games that respond to the player's expressions.\n\n[iOS](/ml-kit/vision/face-detection/ios)\n[Android](/ml-kit/vision/face-detection/android)\n\nKey capabilities\n\n- **Recognize and locate facial features** Get the coordinates of the eyes, ears, cheeks, nose, and mouth of every face detected.\n- **Get the contours of facial features** Get the contours of detected faces and their eyes, eyebrows, lips, and nose.\n- **Recognize facial expressions** Determine whether a person is smiling or has their eyes closed.\n- **Track faces across video frames** Get an identifier for each unique detected face. The identifier is consistent across invocations, so you can perform image manipulation on a particular person in a video stream.\n- **Process video frames in real time** Face detection is performed on the device, and is fast enough to be used in real-time applications, such as video manipulation.\n\nExample results\n\nExample 1\n\nFor each face detected:\n\n| Face 1 of 3 ||\n|---------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| **Bounding polygon** | (884.880004882812, 149.546676635742), (1030.77197265625, 149.546676635742), (1030.77197265625, 329.660278320312), (884.880004882812, 329.660278320312) |\n| **Angles of rotation** | Y: -14.054030418395996, Z: -55.007488250732422 |\n| **Tracking ID** | 2 |\n| **Facial landmarks** | |---------------------|--------------------------------------| | **Left eye** | (945.869323730469, 211.867126464844) | | **Right eye** | (971.579467773438, 247.257247924805) | | **Bottom of mouth** | (907.756591796875, 259.714477539062) | ... etc. |\n| **Feature probabilities** | |--------------------|---------------------| | **Smiling** | 0.88979166746139526 | | **Left eye open** | 0.98635888937860727 | | **Right eye open** | 0.99258323386311531 | |\n\nExample 2 (face contour detection)\n\nWhen you have face contour detection enabled, you also get a list of points\nfor each facial feature that was detected. These points represent the shape of\nthe feature. The following image illustrates how these points map to a face.\nClick the image to enlarge it:\n\n[](/static/ml-kit/vision/face-detection/images/face_contours.svg)\n\n| Facial feature contours ||\n|----------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| **Nose bridge** | (505.149811, 221.201797), (506.987122, 313.285919) |\n| **Left eye** | (404.642029, 232.854431), (408.527283, 231.366623), (413.565796, 229.427856), (421.378296, 226.967682), (432.598755, 225.434143), (442.953064, 226.089508), (453.899811, 228.594818), (461.516418, 232.650467), (465.069580, 235.600845), (462.170410, 236.316147), (456.233643, 236.891602), (446.363922, 237.966888), (435.698914, 238.149323), (424.320740, 237.235168), (416.037720, 236.012115), (409.983459, 234.870300) |\n| **Top of upper lip** | (421.662048, 354.520813), (428.103882, 349.694061), (440.847595, 348.048737), (456.549988, 346.295532), (480.526489, 346.089294), (503.375702, 349.470459), (525.624634, 347.352783), (547.371155, 349.091980), (560.082031, 351.693268), (570.226685, 354.210175), (575.305420, 359.257751) |\n| (etc.) | |"]]
