Intro - TensorFlow Object Detection API Tutorial p.1
Why Stack Overflow recommends it
Stack Overflow users link it as the standard walkthrough for getting the Object Detection API running and for training a custom detector, and they cite it when asking about errors they hit while following along.
Recommended by 5 people across 4 different months, 2017 to 2019.
What it is
Sentdex introduces the TensorFlow Object Detection API in the first part of a tutorial miniseries. It covers detecting objects with bounding boxes in images and video using pre-trained models, with later parts moving to training custom models.
What people said
Recommends Kaushal28 · 2017-11 · Stack Overflow · on “Nothing is being detected in Tensorflow Object detection API”
I'm trying to implement Tensorflow object detection API sample. I am following sentdex [this video] videos for getting started. The sample code runs perfectly, it also shows the images which are used for testing the results, but no boundaries around detected objects are shown. Just the plane image is displayed without any errors. I'm using this code: This Github link [LINK] . This is my result after running the sample code. [LINK] another image without any detection. [LINK] What I'm missing here? The code is included in above link and there is no error logs. Results of box, score, classes, num in that order. [[[ 0.74907303 0.14624023 1. 1. ] [ 0. 0. 1. 1. ] [ 0. 0.20880508 1. 1. ] [ 0.74907303 0.
Recommends cdahms · 2018-01 · Stack Overflow · on “Tensorflow Object Detection API on Windows - error "ModuleNotFoundError: No module named 'utils'"”
I'm attempting to get the TensorFlow Object Detection API [LINK] working on Windows by following the install instructions [LINK] Which seem to be for Linux/Mac. I can only get this to work if I put a script in the directory I cloned the above repo to. If I put the script in any other directory I get this error: ModuleNotFoundError: No module named 'utils' I suspect that the cause is not properly doing the Windows equivalent of this command listed on the install instructions above: # From tensorflow/models/research/ export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim I'm using Windows 10, Python 3.6, and TensorFlow 1.4.0 if that matters. Of course, I've Googled on this concern and found various li
Recommends Sachin Patel · 2018-06 · Stack Overflow · on “How to detect an image Object on tensor-flow?”
It seems that you want to detect a custom object in an image. For custom object here [this video] is a great tutorial. After training clone Tensorflow [LINK] GitHub repository; it has an Android example in the directory tensorflow/tensorflow/examples/android Place your model & label file in assets folder. Open above cloned project in the android studio, it will ask you to sync Gradle; open DetectorActivity.java file; replace the model & label file name with your's in below 2 variable names. TF_OD_API_MODEL_FILE & TF_OD_API_LABELS_FILE Run the android project, on your device 4 apps will get installed from which TF_Detect is your desired app Hope this helps you...
Recommends Aquib · 2018-06 · Stack Overflow · on “Haar Cascade Training for Parts of a Known Object”
Yes, you can train your own object detection like YOLO, Tensorflow object detection API, here is a GitHub repo [LINK] also look in the blog. Here's awesome tutorial series on the same [this video]
Recommends Italo José · 2018-10 · Stack Overflow · on “google colab setting a '^C' in the proccess”
I'm running this code [LINK] that i got from this tutorial [this video] I'm trying running the tensorflow object detection api, all code work well, if you run all calls, all cell will works well, and in the end, my images are classified. Buuut have 1 cell that dont work well, it's work, but doesn't like it must work. When i will train my model with !python legacy/train.py --logtostderr --train_dir=training/ --pipeline_config_path=training/ssd_mobilenet_v1_pets.config it start the tensorflow and start the training, buuut it only run 3 steps, 4 steps, some times 20,21,23 steps and in the end, the google colab set a ^C in the process I never can finish my training because the google colab close my p
Cites bogdan.georgescu.gbm · 2018-01 · Stack Overflow · on “tensorflow object_detection train.py error”
I'm currently trying to retrain Google's Tensorflow Object Detection [LINK] I followed the steps in sentdex's tutorial [this video] (and probably like many of you did). After I generated the tf_records and moved them in the directory models/research/object_detection , everything should be set for the training. To initialize the training process I used: python train.py --logtostderr --train_dir=training/ --pipeline_config_path=training/ssd_mobilenet_v1_pets.config The result should be the command line printing the current step and accuracy reached, but instead i got: Traceback (most recent call last): File "train.py", line 49, in <module> from object_detection import trainer File "/home/ubuntu/ten
Cites Anoop · 2018-04 · Stack Overflow · on “Tensorflow object detection api mis classifying objects”
I followed a simple tutorial [this video] to train a custom object detector. I got my loss up to 0.6, however my issue is that the detected will classify other objects as what I've trained it with. For example in my case it classifies a dog as macarooni and cheese. What am I doing wrong ? [LINK]
Cites Jorge · 2018-05 · Stack Overflow · on “CUDA library not found error with object detection API”
I've been following this tutorial [this video] to run an object detection program from jupyter notebook. When I open the object-detection tutorial, it shows me the following error: could not find 'cudart64_90.dll'. tensorflow requires that this dll be installed in a directory that is named in your %path% environment variable. I've already installed CUDA 9.0 developer kit, and the file 'cudart64_90.dll' exists in my NVIDIA directory. Is the .dll file in the right place? Why am I getting this problem?
Cites alyssaeliyah · 2018-05 · Stack Overflow · on “Add Objects Detected in TensorFlow”
I am currently exploring TensorFlow object detection system, I know it can classify say for example 90 objects as a default. I want to add more objects in the detection. What are the basic steps shall I take? I have found a tutorial in [this video] about tensorflow but it only talks about training specifically new objects not adding more objects to the default objects detected in a certain model classifier. Any kinds of reference are also encouraged. Any help would be very appreciated. I really need this to get working. Thanks.
Cites SineCo · 2018-06 · Stack Overflow · on “ModuleNotFoundError: No module named 'object_detection' while training objects images”
I downloaded Tensorflow object_detection API. I was able to run the tutorial and see the results (Everything went well I followed this video: TensorFlow Object Detection API Tutorial p.1 [this video] ). When I tried to train my own object I got this error: ModuleNotFoundError: No module named 'object_detection' while running this command: python3 generate_tfrecord.py --csv_input=data/train_labels.csv --output_path=data/train.record Note: I ran all these commands but did not work export PYTHONPATH=$PYTHONPATH:`pwd`:`pwd`/slim sudo python3 setup.py build sudo python3 setup.py install sudo pip install -e . Info: I am using Ubuntu 18, Tensorflow 1.8.0