Commit Graph

2773 Commits

Author SHA1 Message Date
Sebastian Schmidt
767db32d69 Support multiple poses for PoseLandmarker
PiperOrigin-RevId: 529430797
2023-05-04 09:29:12 -07:00
Prianka Liz Kariat
3df4f7db64 Updated time out for object detector 2023-05-04 20:00:29 +05:30
Prianka Liz Kariat
08282d9fd7 Updated time out for image classifier async tests 2023-05-04 19:59:39 +05:30
Prianka Liz Kariat
a4e11eac78 Added constants for time out 2023-05-04 19:55:19 +05:30
Prianka Liz Kariat
00712d727e Updated wait time for object detector tests 2023-05-04 19:53:06 +05:30
Prianka Liz Kariat
33ae23c53a Increased wait time for image classifier asynchronous tests 2023-05-04 19:51:54 +05:30
Prianka Liz Kariat
8ec0724b65 Updated documentation to include note about rgba images 2023-05-04 19:44:11 +05:30
Prianka Liz Kariat
ddd1515f88 Updated documentation 2023-05-04 19:40:15 +05:30
Prianka Liz Kariat
d401439daa Updated formatting 2023-05-04 19:22:11 +05:30
Prianka Liz Kariat
1136d4d515 Updated CVPixelBuffer to support pixel format type of 32RGBA 2023-05-04 18:58:49 +05:30
Prianka Liz Kariat
7a7f27c34b Merge branch 'ios-normalized-keypoint-hash' into ios-object-detector-async-fixes 2023-05-04 17:21:36 +05:30
Prianka Liz Kariat
381ffcb474 Added hash implementation for iOS normalized keypoint 2023-05-04 17:10:07 +05:30
Prianka Liz Kariat
87593a2ade Updated docuemntation of MPPObjectDetector 2023-05-04 17:05:17 +05:30
Prianka Liz Kariat
ab135190e5 Updated iOS object detector to use delegates instead of callbacks for async calls 2023-05-04 17:03:40 +05:30
Prianka Liz Kariat
a21c08bf4d Added method for creating unique dispatch queue names in MPPVisionTaskRunner 2023-05-04 17:00:12 +05:30
Prianka Liz Kariat
e47bb16544 Added validation of C++ image classification result packet in MPPImageClassifierResult+Helpers.mm 2023-05-04 16:52:58 +05:30
Prianka Liz Kariat
ab4b07646c Updated MPPImageClassifier to use delegates instead of completion blocks for callback. 2023-05-04 16:43:18 +05:30
Prianka Liz Kariat
1323a5271c Added method to create unique dispatch queue names in MPPVisionTaskRunner 2023-05-04 16:39:43 +05:30
Jiuqiang Tang
c6e3f08282 Expose FaceAligner and LanguageDetector to be public MediaPipe Tasks Python API.
PiperOrigin-RevId: 529227382
2023-05-03 16:38:49 -07:00
MediaPipe Team
b350f72394 Support MultiHW AVG Architecture for object detector
PiperOrigin-RevId: 529221127
2023-05-03 16:12:59 -07:00
Copybara-Service
8c324fbd77 Merge pull request #4325 from kinaryml:language-detector-python
PiperOrigin-RevId: 529213966
2023-05-03 15:50:42 -07:00
MediaPipe Team
a09e39d431 Add TransformerParameters proto
PiperOrigin-RevId: 529213840
2023-05-03 15:46:50 -07:00
Yuqi Li
e428bdb7e8 internal change.
PiperOrigin-RevId: 529181374
2023-05-03 13:37:16 -07:00
MediaPipe Team
606b83ac65 Internal change
PiperOrigin-RevId: 529180655
2023-05-03 13:32:44 -07:00
MediaPipe Team
7c955246aa Support scribble input for Interactive Segmenter Java API
PiperOrigin-RevId: 529177660
2023-05-03 13:22:09 -07:00
Copybara-Service
e84e90e5b2 Merge pull request #4361 from kinaryml:face-aligner-python
PiperOrigin-RevId: 529165597
2023-05-03 12:39:23 -07:00
MediaPipe Team
c780559214 Internal change
PiperOrigin-RevId: 529161249
2023-05-03 12:20:44 -07:00
MediaPipe Team
09662749ea Support scribble input for Interactive Segmenter
PiperOrigin-RevId: 529156049
2023-05-03 12:01:10 -07:00
MediaPipe Team
baa8fc68a1 Make uploading to GPU optional in Image.GetGpuBuffer().
PiperOrigin-RevId: 529066617
2023-05-03 05:57:58 -07:00
MediaPipe Team
3789156a41 Internal change
PiperOrigin-RevId: 529011480
2023-05-03 00:29:13 -07:00
MediaPipe Team
1dea01aecc Internal change
PiperOrigin-RevId: 528996603
2023-05-02 22:57:38 -07:00
MediaPipe Team
c698381e48 Internal change
PiperOrigin-RevId: 528939095
2023-05-02 18:05:31 -07:00
Jiuqiang Tang
bf11fb313e Expose PoseLandmarker as a public MediaPipe Tasks Python API.
PiperOrigin-RevId: 528882303
2023-05-02 14:04:01 -07:00
Jiuqiang Tang
4d9812af43 Pose detector uses advanced_gpu_api for gpu inference to resolve unsupported gpu op issue.
PiperOrigin-RevId: 528879218
2023-05-02 13:52:26 -07:00
Yuqi Li
9ce16fddeb nit: format the documentation of LandmarksDetectionResult.
PiperOrigin-RevId: 528848566
2023-05-02 11:56:47 -07:00
MediaPipe Team
421c9e8e97 Fix typo
PiperOrigin-RevId: 528829423
2023-05-02 10:51:47 -07:00
MediaPipe Team
4d112c132f Fix msan errors.
PiperOrigin-RevId: 528825081
2023-05-02 10:37:50 -07:00
Jiuqiang Tang
60055f6fee Add more comments and usage example of the face stylizer graph.
PiperOrigin-RevId: 528823127
2023-05-02 10:32:43 -07:00
MediaPipe Team
7fdbbee5be Internal change
PiperOrigin-RevId: 528799585
2023-05-02 09:07:29 -07:00
MediaPipe Team
5b93477589 internal change
PiperOrigin-RevId: 528719459
2023-05-02 02:17:18 -07:00
Chuo-Ling Chang
3719aaef7e Fix typo.
PiperOrigin-RevId: 528693117
2023-05-01 23:52:45 -07:00
MediaPipe Team
0a8be0d09d Internal change
PiperOrigin-RevId: 528632873
2023-05-01 18:40:33 -07:00
Jiuqiang Tang
fca728d226 Set face alignment image width and hight to 256.
PiperOrigin-RevId: 528583074
2023-05-01 15:01:39 -07:00
Jiuqiang Tang
162a999887 Check the output stream tag rather than the input stream tag in face stylizer graph.
PiperOrigin-RevId: 528555024
2023-05-01 13:14:32 -07:00
MediaPipe Team
5526e96b21 Internal change for proto library outputs.
PiperOrigin-RevId: 528539840
2023-05-01 12:17:29 -07:00
Yuqi Li
085f8265fb Internal change
PiperOrigin-RevId: 528517562
2023-05-01 11:01:23 -07:00
Kinar R
544e4b66f7
Merge branch 'google:master' into face-aligner-python 2023-05-01 22:56:00 +05:30
MediaPipe Team
cab619f8da Fix typo in README
PiperOrigin-RevId: 528506206
2023-05-01 10:25:05 -07:00
Jiuqiang Tang
b9a9da5de5 Ignore fetching face stylizer model when the graph doesn't output stylized face images.
PiperOrigin-RevId: 528504312
2023-05-01 10:18:41 -07:00
kinaryml
bd039f8b65 Updated necessary BUILD files 2023-05-01 05:56:52 -07:00