So I faced the issue where only the first frame is detected and tracked when I running the object_tracker.py using tensorflow-gpu. However, once I uninstalled tensorflow-gpu and replaced it with tensorflow-cpu (latest version since I am using python 3.9), it works flawlessly. Does anyone one faced this issue? I am using GTX 1050 on my laptop which I believe it should run since emasterclassacademy (an author who ran this code on YOUTUBE with GTX1050ti) was able to run the code . FYI, I am running on latest CUDA version 11 as well. Below is the full GPU information I have gotten when I ran with tensorflow-gpu.
So I faced the issue where only the first frame is detected and tracked when I running the object_tracker.py using tensorflow-gpu. However, once I uninstalled tensorflow-gpu and replaced it with tensorflow-cpu (latest version since I am using python 3.9), it works flawlessly. Does anyone one faced this issue? I am using GTX 1050 on my laptop which I believe it should run since emasterclassacademy (an author who ran this code on YOUTUBE with GTX1050ti) was able to run the code . FYI, I am running on latest CUDA version 11 as well. Below is the full GPU information I have gotten when I ran with tensorflow-gpu.