tensorflow / tensorflow/tflite-support

TfLiteGpuDelegate Prepare: Output size is less than input size.

Open
#935 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
C++
Stars
441
Forks
146
PR merge metrics
No merged PRs in 30d

Description

I am inferencing a float model which has input size of 1x480x640x3 and output size of 1x480x640x3,
I am building the AAR in the mediapipe gogole for inferencing and using that AAR in the android studio.
GPU delegate is failed to prepare saying the error, output size is less than the input size at number 34(which last layer of the tflite model).
Following is the error
TfLiteGpuDelegate Prepare: Output size is less than input size.
Node number 34 (TfLiteGpuDelegate) failed to prepare.
Screenshot from 2023-08-04 16-28-19
Screenshot from 2023-08-04 16-28-19

MediaPipe Face Mesh error:MediaPipe packet creation error: internal: Graph has errors:
Calculator::Open() for node "llefacerecgpu__inferencecalculator__llefacerecgpu__InferenceCalculator" failed: ; (interpreter_->ModifyGraphWithDelegate(delegate_.get()))==(kTfLiteOk)gl.cc:456)

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing the Android AAR inference with the reported 1x480x640x3 model and GPU delegate. Trace the MediaPipe InferenceCalculator failure through ModifyGraphWithDelegate and the reported node 34 error; no source file or test is named, so completion would require identifying the delegate/model incompatibility and documenting or fixing the cause.

Written by the indexing model from the issue text.

Assessment

Tech stack
android, cpp
Domain
machine-learning, mobile-dev
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.