Following the code here: https://github.com/tensorflow/models/blob/master/research/object_detection/object_detection_tutorial.ipynb
No matter the image inputted, there seems to be a hard limit of 20 objects detected. Example:
The problem is also seen in this post: TensorFlow Object Detection API Weird Behavior
Is there some configuration or parameter that can be changed to raise the number of objects detected?
EDIT: I can confirm that greater than 20 objects are detected, but there is a maximum of 20 that will be shown in the final output. Is there a way to increase this limit?
The max number of detections can be set in your config file. By default it's usually 300, so you should be fine.
Your problem here is the number of displayed detections. Towards the end of your code you have a call to vis_util.visualize_boxes_and_labels_on_image_array. Just add max_boxes_to_draw=None to its arguments to display all the detections (or choose some bigger number if you want).
Related
I'm working in a co-simulation project between Simulink and Gazebo. The aim is to move a robot model in Gazebo with the trajectory coordinates computed from Simulink. I'm using MATLAB R2022a, ROS 2 Dashing and Gazebo 9.9.0 in a computer running Ubuntu 18.04.
The problem is that when launching the FMU with the fmi_adapter, I'm obtaining the following. It is tagged as [INFO], but actually messing up all my project.
[fmi_adapter_node-1] [INFO] [fmi_adapter_node]: Simulation time 1652274762.959713 is greater than timer's time 1652274762.901340. Is your step size to large?
Note the timer's time is higher than the simulation time. Even if I try to change the step size with the optional argument of the fmi_adapter_node, the same log appears with small differences in the times. I'm using the next commands:
ros2 launch fmi_adapter fmi_adapter_node.launch.py fmu_path:=FMI/Trajectory/RobotMARA_SimulinkFMU_v2.fmu # default step size: 0.2
ros2 launch fmi_adapter fmi_adapter_node.launch.py fmu_path:=FMI/Trajectory/RobotMARA_SimulinkFMU_v2.fmu _step_size:=0.001
As you would expect, the outputs of the FMU are the xyz coordinates of the robot trajectory in each time step. Since the fmi_adapter_node creates topics for both inputs and outputs, I'm reading the output xyz values by means of 3 subscribers with the next code. Then, those coordinates are being used to program the robot trajectories with the MoveIt-Python API.
When I run the previous Python code, I'm obtaining the following warning once and again and the robot manipulator actually doesn't move.
[ WARN] [1652274804.119514250]: TF_OLD_DATA ignoring data from the past for frame motor6_link at time 870.266 according to authority unknown_publisher
Possible reasons are listed at http://wiki.ros.org/tf/Errors%20explained
The previous warning is explained here, but I'm not able to fix it. I've tried clicking Reset in RViz, but nothing changes. I've also tried the following without success:
ros2 param set /fmi_adapter_node use_sim_time true # it just sets the timer's time to 0
It seems that the clock is taking negative values, so there is a synchronization problem.
Any help is welcome.
The warning message by the FMIAdapterNode is emitted if the timer's period is only slightly greater than the simulation step-size and if the timer is preempted by other processes or threads.
I created an issue at https://github.com/boschresearch/fmi_adapter/issues/9 which explains this in more detail and lists two possible fixes. It would be great if you could contribute to this discussion.
I assume that the TF_OLD_DATA error is not related to the fmi_adapter. Looking at the code snippet at ROS Answers, I wondered whether x,y,z values are re-published at all given that the lines
pose.position.x = listener_x.value
pose.position.y = listener_y.value
pose.position.z = listener_z.value
are not inside a callback and executed even before rospy.spin(), but maybe that's just truncated.
I wrote a Python script that tends to crash sometimes with a Memory Allocation Error. I noticed that the pagefile.sys of my Win10 64 system skyrockets in this script and exceeds the free memory.
My current solution is to run the script in steps, so that every time the script runs through, the pagefile empties.
I would like the script to run through all at once, though.
Moving the pagefile to another drive is not an option, unfortunately, because I only have this one drive and moving the pagefile to an external drive does not seem to work.
During my research, I found out about the module gc but that is not working:
import gc
and after every iteration I use
gc.collect()
Am I using it wrong or is there another (python-based!) option?
[Edit:]
The script is very basic and only iterates over image files (using Pillow). The script only checks for width, height and resolution of the image, calculates the dimensions in cm.
If height > width, the image is rotated 90° counterclockwise.
The images are meant to be enlarged or shrunk to A3 size (42 x 29.7cm), so I use the width/height ratio to calculate whether I can enlarge the width to 42cm and the height remains < 29.7cm and in case the height is > 29.7cm, I enlarge the height to 29.7 cm.
For the moment, I do the actual enlargement/shrinking still in Photoshop. Based on whether it is a width/height enlargement, the file is moved to a certain folder that contains either one of those file types.
Anyways, the memory explosion happens in the iteration that only reads the file dimensions.
For that I use
with Image.open(imgOri) as pic:
widthPX = pic.size[0]
heightPX = pic.size[1]
resolution = pic.info["dpi"][0]
widthCM = float(widthPX) / resolution * 2.54
heightCM = float(heightPX) / resolution * 2.54
I also calculate whether the shrinking would be too strong, the image gets divided in half and re-evaluated.
Even though it is unnecessary, I still added pic.close
to the with open()statement, because I thought Python may be keeping the image files open, but that didn't help.
Once the iteration finished, the pagefile.sys goes back to its original size, so in case that error occurs, I take some files out and do them gradually.
So im using this Node module to connect to chef from my API.
https://github.com/normanjoyner/chef-api
The same contains a method called "partialSearch" which will fetch determined data for all nodes that match a given criteria. The problem I have, on of our environments have 1386 nodes attached to it, but it seems the module only returns 1000 as a maximum.
There does not seem to be any method to "offset" the results. This module works pretty well and its a shame this feature is not implemented since its lack really breaks the utility of such.
Does someone bumped into a similar issue with this module and can advise how to workaround it?
Here its an extract of my code :
chef.config(SetOptions(environment));
console.log("About to search for any servers ...");
chef.partialSearch('node',
{
q: "name:*"
},
{
name: ['name'] ,
'ipaddress': ['ipaddress'] ,
'chef_environment': ['chef_environment'] ,
'ip6address': ['ip6address'],
'run_list': ['run_list'],
'chef_client': ['chef_client'],
'ohai_time': ['ohai_time']
}
, function(err, chefRes) {
Regards!
The maximum is 1000 results per page, you can still request pages in order. The Chef API doesn't have a formal cursor system for pagination so it's just separate requests with a different start value, which can sometimes lead to minor desync (as in an item at the end of one page might shift in ordering and also show up at the start of the next page) so just make sure you handle that. That said, the fancy API in the client library you linked doesn't seem to expose that option, so you'll have to add it or otherwise workaround the problem. Check out https://github.com/sethvargo/chef-api/blob/master/lib/chef-api/resources/partial_search.rb#L34 for a Ruby implementation that does handle it.
We have run into similar issues with Chef libraries. One work-around you might find useful is if you have some node attribute that you can use to segment all of your nodes into smaller groups that are less than 1000.
If you have no such natural segmentation friendly already, a simple implementation would be to create a new attribute called segment and during your chef runs set the attribute's value randomly to a number between 1 and 5.
Now you can perform 5 queries (each query will only search for a single segment) and you should find all your nodes and if the randomness is working each group will be sized about 275 (1386/5).
As your node population grows you'll need to keep increasing the number of segments to ensure the segment sizes are less than 1000.
I am writing a custom control which inherits from FrameworkElement. I use DrawingVisual to render to the screen. Therefore, I have to calculate myself the optimal size of the Visuals I draw. If HorizontalAlignment is set to Stretch, I would like to use all available space. Only WPF doesn't tell how much space is really available :-(
In MeasureOverride, the constraint.Width is Infinity. I need to return a desired size which is less than Infinity, otherwise I get the exception "InvalidOperationException: Layout measurement override of element 'xyz' should not return PositiveInfinity as its DesiredSize, even if Infinity is passed in as available size."
I tried to return 0 in the hope that Arrange would pass all the available space this time, since in MeasureOverride it said Infinity width is available. But ArrangeOverride says the arrangeBounds.Width is 0.
I tried to return an arbitrary width (well, actually I returned the Monitor width to MeasureOverride), but then ArrangeOverride tells me I can use the complete monitor, which is, of course, not true :-(
So which width do I return in MeasureOverride to get the maximum space really available ?
Update 23.1.23
My control is in a library and has no idea who is the container holding it. I inherit from that control to make many different controls, for example a line graph that can contain millions of measurements. So if there is unlimited space, the graph gets millions of pixel wide. If the space is limited, let's say 1000 pixels, then the graph displays in 1 pixel the average of thousands of measurements.
After using WPF now for nearly 20 years, I think the proper solution is to raise an exception when the available width is Infinity, but my control cannot deal with that. The exception demands that the control is put into a container which gives limited width to its children.
For example a ScrollViewer gives unlimited space to its children. But since my line graph, or any other control I design that can not deal with unlimited space, gets unreasonably big, it is better to alert the developer using the control immediately that the control needs to be hosted in a container like a Grid, where the Grid tells the child the exact size available. Of course, also the Grid might give unlimited space when the column width is set to auto, but in this case my Control should raise an exception, so that the column width can get changed to a number of pixels or *.
Everyone knows that MRTG needs at least one value to be passed on it's input.
In per-target options MRTG has 'gauge', 'absolute' and default (with no options) behavior of 'what to do with incoming data'. Or, how to count it.
Lets look at the elementary, yet popular example :
We pass cumulative data from network interface statistics of 'how much packets were recieved by the interface'.
We take it from '/proc/net/dev' or look at 'ifconfig' output for certain network interface. The number of recieved bytes is increasing every time. Its cumulative.
So as i can imagine there could be two types of possible statistics:
1. How fast this value changes upon the time interval. In oher words - activity.
2. Simple, as-is growing graphic that just draw every new value per every minute (or any other time interwal)
First graphic will be saltatory (activity). Second will just grow up every time.
I read twice rrdtool's and MRTG's docs and can't understand which option mentioned above counts what.
I suppose (i am not sure) that 'gauge' draw values as is, without any differentiation calculations (good for measuring how much memory or cpu is used every 5 minutes). And default or 'absolute' behavior tryes to calculate the speed between nearby measures, but what's the differencr between last two?
Can you, guys, explain in a simple manner which behavior stands after which option of three options possible?
Thanks in advance.
MRTG assumes that everything is being measured as a rate (even if it isnt a rate)
Type 'gauge' assumes that you have already calculated the rate; thus, the provided value is stored as-is (after Data Normalisation). This is appropriate for things like CPU usage.
Type 'absolute' assumes the value passed is the count since the last update. Thus, the value is divided by the number of seconds since the last update to get a rate in thingies per second. This is rarely used, and only for certain unusual data sources that reset their value on being read - eg, a script that counts the number of lines in a log file, then truncates the log file.
Type 'counter' (the default) assumes the value passed is a constantly growing count, possibly that wraps around at 16 or 64 bits. The difference between the value and its previous value is divided by the number of seconds since the last update to get a rate in thingies per second. If it sees the value decrease, it will assume a counter wraparound at 16 or 64 bit. This is appropriate for something like network traffic counters, which is why it is the default behaviour (MRTG was originally written for network traffic graphs)
Type 'derive' is like 'counter', but will allow the counter to decrease (resulting in a negative rate). This is not possible directly in MRTG but you can manually create the necessary RRD if you want.
All types subsequently perform Data Normalisation to adjust the timestamp to a multiple of the Interval. This will be more noticeable for Gauge types where the value is small than for counter types where the value is large.
For information on this, see Alex van der Bogaerdt's excellent tutorial