Groups

MediaPipe

Contact owners and managers

1–30 of 277

Welcome to the discussion forum for MediaPipe, a cross platform framework for building multimodal (eg. video, audio, any time series data) applied ML pipelines.

This forum is for general discussion on MediaPipe (https://mediapipe.dev). Any issues with the forum, pls email mediapipe@google.com

0 selected

12/23/25

License for Model bundles (e.g. face_landmarker.task ?)

Hello everyone, Does anyone know if the model bundles (like the FaceLandmarker found here) is

unread,

License for Model bundles (e.g. face_landmarker.task ?)

Hello everyone, Does anyone know if the model bundles (like the FaceLandmarker found here) is

12/23/25

11/25/25

Face mesh landmark picker web app

I had to pick landmarks I was interested in, but I couldn't find anything less than "

unread,

Face mesh landmark picker web app

I had to pick landmarks I was interested in, but I couldn't find anything less than "

11/25/25

11/21/25

Is it able to modify existed .task?

Hi guys, I'm studying & trying to use mediapipe in my Android Kotlin project I want to modify

unread,

Is it able to modify existed .task?

Hi guys, I'm studying & trying to use mediapipe in my Android Kotlin project I want to modify

11/21/25

10/1/25

PostDoc fellowship opportunity from DigitalFuture in Stockholm!

Are you (or someone you know) looking for a postdoc fellowship that combines motion track, AI, brain

unread,

PostDoc fellowship opportunity from DigitalFuture in Stockholm!

Are you (or someone you know) looking for a postdoc fellowship that combines motion track, AI, brain

10/1/25

Andrea Fiorito, Federica Paolì4

9/22/25

mediapipe hand tracking on Android, C++ ndk side

Hi, do you still need this? On Friday, September 5, 2025 at 5:56:13 PM UTC+1 Federica Paolì wrote: Hi

unread,

mediapipe hand tracking on Android, C++ ndk side

Hi, do you still need this? On Friday, September 5, 2025 at 5:56:13 PM UTC+1 Federica Paolì wrote: Hi

9/22/25

28.Lê Trung Hiếu

9/8/25

Why use Butterworth filter after MediaPipe smoothing (One-Euro) for pose landmarks?

I'm working on human pose estimation using **MediaPipe** (Pose or Face Mesh), which applies a **

unread,

Why use Butterworth filter after MediaPipe smoothing (One-Euro) for pose landmarks?

I'm working on human pose estimation using **MediaPipe** (Pose or Face Mesh), which applies a **

9/8/25

Inbal V, Swappy Asabe2

8/17/25

help with autoflip in colab

very simple On Monday, 5 May 2025 at 01:26:37 UTC+5:30 Inbal V wrote: Hi team, Is there any chance to

unread,

help with autoflip in colab

very simple On Monday, 5 May 2025 at 01:26:37 UTC+5:30 Inbal V wrote: Hi team, Is there any chance to

8/17/25

8/1/25

Scope of dependencies for MCP server

I'm using the mediapipe text embedder to do a very simple retrieval based on embedding dot

unread,

Scope of dependencies for MCP server

I'm using the mediapipe text embedder to do a very simple retrieval based on embedding dot

8/1/25

7/29/25

how can i get the number of how many times person mouth open/close in video

unread,

how can i get the number of how many times person mouth open/close in video

7/29/25

Venkat Raghavulu, … Haileyesus3

7/23/25

Mediapipe LLM build fails for Android X86_64

Hey Venkat, I'm also facing the same problem. My emulator device is X86_64. Is there anything you

unread,

Mediapipe LLM build fails for Android X86_64

Hey Venkat, I'm also facing the same problem. My emulator device is X86_64. Is there anything you

7/23/25

6/28/25

Mediapipe: processing landmark coordinates

Hi, I'm trying to figure out ways to process landmark coordinates generated with pose estimation

unread,

Mediapipe: processing landmark coordinates

Hi, I'm trying to figure out ways to process landmark coordinates generated with pose estimation

6/28/25

Trọng Nguyễn Công

6/10/25

Xcode: Duplicate Symbols (Conflicts with TensorFlowLiteSelectTfOps dependencies)

I'm also running into this issue when integrating MediaPipeTasksGenAIC together with

unread,

Xcode: Duplicate Symbols (Conflicts with TensorFlowLiteSelectTfOps dependencies)

I'm also running into this issue when integrating MediaPipeTasksGenAIC together with

6/10/25

5/26/25

for double player

any methods to stabilize the output when numPoses >= 2, wai yuk在 2025年5月26日星期一下午8:49:16 [UTC+8]寫道：

unread,

for double player

any methods to stabilize the output when numPoses >= 2, wai yuk在 2025年5月26日星期一下午8:49:16 [UTC+8]寫道：

5/26/25

Muhammad Afshal

5/20/25

Rotation-Invariant Distance Between 3D Face Landmarks in Browser Using @mediapipe/tasks-vision

Hi Team, I'm currently working on a browser-based application using the @mediapipe/tasks-vision

unread,

Rotation-Invariant Distance Between 3D Face Landmarks in Browser Using @mediapipe/tasks-vision

Hi Team, I'm currently working on a browser-based application using the @mediapipe/tasks-vision

5/20/25

5/5/25

Frame Rate Enquiry

Hi All, I have implemented custom object detection and pose detection on the web using

unread,

Frame Rate Enquiry

Hi All, I have implemented custom object detection and pose detection on the web using

5/5/25

4/14/25

I'm trying to use the facial landmarks that the mediapipe provides. The only problem I have is

unread,

I'm trying to use the facial landmarks that the mediapipe provides. The only problem I have is

4/14/25

4/5/25

I wanna to detect these handlandkmarker gesture but i don't know how.

I don't know how to differentiate them by using the handlandmark but I don't know how 在 2025年

unread,

I wanna to detect these handlandkmarker gesture but i don't know how.

I don't know how to differentiate them by using the handlandmark but I don't know how 在 2025年

4/5/25

4/5/25

I wanna to detect these handlandkmarker gesture but i don't know how.

unread,

I wanna to detect these handlandkmarker gesture but i don't know how.

4/5/25

Eric Bubar, Niko Avradopoulos2

3/28/25

Drawing landmarks on live webcam feed

+1 looking for a migrated draw landmarks solution! On Thursday, May 11, 2023 at 7:08:17 PM UTC-4 ebu.

unread,

Drawing landmarks on live webcam feed

+1 looking for a migrated draw landmarks solution! On Thursday, May 11, 2023 at 7:08:17 PM UTC-4 ebu.

3/28/25

2/14/25

MediaPipe GPU Backend - Entirely different results to CPU Backend

Hi there, I've been training my own image segmenter model that operates on 512x512 input, similar

unread,

MediaPipe GPU Backend - Entirely different results to CPU Backend

Hi there, I've been training my own image segmenter model that operates on 512x512 input, similar

2/14/25

2/13/25

Issues with Image Orientation Handling in AR App (MediaPipe + YOLO on Android)

Hi everyone, I'm working on an AR application using the MediaPipe plugin for image processing and

unread,

Issues with Image Orientation Handling in AR App (MediaPipe + YOLO on Android)

Hi everyone, I'm working on an AR application using the MediaPipe plugin for image processing and

2/13/25

Heidi He, Max Chan3

1/27/25

Questions About MediaPipe Privacy Policy

I think I replied but maybe the response is not sent correctly so it is not shown in the discussion?

unread,

Questions About MediaPipe Privacy Policy

I think I replied but maybe the response is not sent correctly so it is not shown in the discussion?

1/27/25

menashe soffer, … יפי קולפ5

1/27/25

face mesh depth coordinate

האינדקסים של הנקודות נשאר תמיד אותו הדבר ורק הערכים של הנקודות משתנה למשל הנקודה הראשונה מייצגת תמיד

unread,

face mesh depth coordinate

האינדקסים של הנקודות נשאר תמיד אותו הדבר ורק הערכים של הנקודות משתנה למשל הנקודה הראשונה מייצגת תמיד

1/27/25

Jason Williams, … Max Chan4

1/14/25

Building for web

Thank you! Hope you can find ur solution in other places. May I ask where else did you ask? I can

unread,

Building for web

Thank you! Hope you can find ur solution in other places. May I ask where else did you ask? I can

1/14/25

1/6/25

Tracking module part of face detection in video?

Hello team, I'm interested in using face detection on pre-recorded videos (not live), and I'd

unread,

Tracking module part of face detection in video?

Hello team, I'm interested in using face detection on pre-recorded videos (not live), and I'd

1/6/25

11/30/24

Fastest inference backend for MediaPipe models on Raspberry Pi board

Hi all! Me and my team is building an AI inference backend, especially focusing on CPU optimization

unread,

Fastest inference backend for MediaPipe models on Raspberry Pi board

Hi all! Me and my team is building an AI inference backend, especially focusing on CPU optimization

11/30/24

11/22/24

Getting issue while integrating mediaipe in flutter app

I am trying to integrate mediapipe poselandmarker in flutter app with the help of method channel. The

unread,

Getting issue while integrating mediaipe in flutter app

I am trying to integrate mediapipe poselandmarker in flutter app with the help of method channel. The

11/22/24

Miracle Okporiaku

11/19/24

Should we resize training image before training to Mediapipe model maker

I've seen various articles that says that resizing all images to a specific aspect ratio or size

unread,

Should we resize training image before training to Mediapipe model maker

I've seen various articles that says that resizing all images to a specific aspect ratio or size

11/19/24

Oddbjørn Overskott, Andrea Fiorito2

11/15/24

Multithreading possible in pose detection?

Hi Oddbjørn, MediaPipe is composed of two main sections, Solutions and Framework. I guess you're

unread,

Multithreading possible in pose detection?

Hi Oddbjørn, MediaPipe is composed of two main sections, Solutions and Framework. I guess you're

11/15/24

10/16/24

IDE to develop Mediapipe Framework?

As per title, what are the most common IDEs or equivalents to develop Mediapipe? Because using vim as

unread,

IDE to develop Mediapipe Framework?

As per title, what are the most common IDEs or equivalents to develop Mediapipe? Because using vim as

10/16/24

Search

Clear search

Close search

Google apps

Main menu