esimov / pigo Goto Github PK

there is two image with same person in it, can this library find and compare face from that two image ?, thankyou

I'm using pigo in my realtime app and observing many false detections. How can i reduce amount of them? Does pigo have some parameters to adjust?

can I share it between go routines?

hard to compile python code.
can youwrite a go demo,not use python and go.
thanks

Can this be somehow used to smart-detect background image and replace it in realtime?

Would you be able to explain how you went and generated the facial landmark / density tree files that are used in the implementation of the Fast Localization of Facial Landmark Points? I would be keen to contribute, or ask if you can, the jaw line to the list of supported facial landmarks

Hi there,
I try to exec demo with Python 3.9 on Windows, sometimes got unexpected exit, with code -1073741819 (0xC0000005) , that err from C shared lib I think, and that because of index out of range usually.
I'm not good at Golang and not founded the true reason from code, give me some advice to fix it please, thx.

Cannot fetch the cascade files on the Webassembly demo when the github server is down raw.githubusercontent.com is down.

Webcam can be used from Go with GoCV. Resulting mat can then be encoded to jpg if needed.

# github.com/esimov/pigo/webcam
src/github.com/esimov/pigo/webcam/main.go:127:28: multiple-value pigo.Unpack() in single-value context

Found by smallrepo.

Build log: https://smallrepo.com/builds/20180607-063653-c1d35e4f

Like this image

It processed but can not open

For some reason the x isn't really x but it's the y value, and vice versa. Weirdly they seem backwards in the json output.

I tested this by opening the image in photoshop and seeing where the x / y sits as compared to the values reported by pigo. They are accurate, just backwards. Unless I've clearly done something nooby 😕

Thanks for sharing this awesome project!

Is it possible to get bounding box coordinates outputted to the command line? I'd like to use pigo on my raspberry pi 3 to automatically crop images captured by the raspberry pi camera.

Also, would this work with people who are wearing glasses?

Many thanks!

Bug with Webassembly (WASM) version:

$ go version
go version go1.16.6 darwin/amd64

Running the Wasm demo after certain time the browser crashes because of Out of memory error. Checking the
memory consumption this increases rapidly.

Screenshots

Additional context

Hello, I had a bug when I execute demo. It's nothing bad ;) .

github.com/esimov/pigo/examples/facedet # python2 demo.py
Traceback (most recent call last):
File "demo.py", line 66, in
cv2.circle(frame, (int(det[1]), int(det[0])), int(det[2]/2.0), (0, 0, 255), 2)
OverflowError: signed integer is greater than maximum

Hi. I'm trying to implement webcam based facial motion capture with with library.

The current set of face and retina landmarks is quite good, but I'm wondering if it's possible to extend the landmark regions that the PICO process can identify.

Specifically:

Option 1: adding new face landmarks

I'm assuming that new facial landmark locations would require defining new cascade files. Did the paper authors mentioned any recommendations on collecting and tagging example training data, or the training process itself?

Option 2: adding simple paint based landmarks

Perhaps a simpler option would be to somehow define cascades for recognizing simple shapes like dot marks, or cross marks placed in an actor's face where the missing facial landmark points are needed.

Let me know if I missed anything.

Thanks!

Resources

AutoDesk SoftImage - The Motion Capture Process

First Tks for your nice project.
I am a bit confuse of the mean of iou parameter and it impact the result of the detection, and i use this lib for face detection, I'm wonder of how to set iouThreshold value correctly.

I was wondering why func Unpack() (*Pigo, error) needs to be a method of the struct Pigo if that struct is never used inside the method. I think making Unpack() a method of the package, instead of the struct, would have less overhead when initializing the classifier.

What do you think?

Is it possible to use oval shape instead of square or circle?

my demo is : https://github.com/Taoja/web-wasm-faceDetector

there have two problems:

• the func rgbaToGrayscale is useless, the picture still colorful when i output imageData and use canvas render it
• unable detect face, maybe it is because func rgbaToGrayscale is useless

Describe the bug

-fplc flag doesn't seem to work if -plc isn't specified.

This command works fine:
$ pigo -in image.jpg -out output.jpg -cf ../pigo/cascade/facefinder -plc ../pigo/cascade/puploc -flpc ../pigo/cascade/lps

But running the following only detects the face but not its facial landmarks:
$ pigo -in image.jpg -out output.jpg -cf ../pigo/cascade/facefinder -flpc ../pigo/cascade/lps

Expected behavior

-flpc should work without -plc as per the screenshot https://user-images.githubusercontent.com/883386/66802771-3b0cc880-ef26-11e9-9ee3-7e9e981ef3f7.png

Would it be possible to extend this library or use it as-is to perform facial recognition and thus identity of persons in images? (as opposed to face detection)

Hi there, is it possible to print/output the Facial Coordinates/Pupil Localization pointers in the JSON file with -json itself? Or any work around for the same?

or any other way to print the exact coordinates of the pupils or the other facial features?

$> ./pigo -in input.jpg -out output.jpg -cf data/facefinder
Processing... -panic: runtime error: index out of range

goroutine 1 [running]:
github.com/esimov/pigo/core.(*Pigo).Unpack(0xc4206a8d50, 0xc420672000, 0x0, 0x200, 0x200, 0x0, 0x0)
        /home/esimov/Projects/Go/src/github.com/esimov/pigo/core/pigo.go:66 +0xc5e
main.(*faceDetector).detectFaces(0xc4206a8f40, 0x7ffe8b90d545, 0x2d, 0x3, 0x3, 0x1, 0x50a920, 0xc42007e0c0)
        /home/esimov/Projects/Go/src/github.com/esimov/pigo/cmd/pigo/main.go:164 +0x23b
main.main()
        /home/esimov/Projects/Go/src/github.com/esimov/pigo/cmd/pigo/main.go:95 +0x296

pigo v1.1.0

What am I missing?

This particular line is a possible (an in fact, experienced) infinite loop:

I've figured it out from the stack trace:

github.com/esimov/pigo/core.(*Pigo).classifyRegion(0xc0001a40a0, 0x42, 0x36, 0x9, 0xc00068e000, 0x238c, 0x238c, 0x5b, 0xbf800000)

The gorutine track trace parameters are as follows:

• 0xc0001a40a0 (the receiver)
• 0x42 row (constantly modified)
• 0x36 col (constantly modified)
• 0x9 scale (!!does not change!!)
• other params (do not change)

The issue comes from a rounding /casting error. The main readme (as well as my code) sets the cascade parameters scale factor as 1.1. When the scale is 9, the factor would come up with 9.9, which again becomes 9 because of the int() conversion.

The particular issue is that the source image was 91x100 (don't even ask), and I've calculated the face min/max size as 10% and 50% of the image width respectively. The minimum face size thus becomes 9 and with the scale ratio of 1.1 = infinite loop.

Possible fixes:

• Clamp the minimum value for a face size where it wouldn't cause problems (10px)
• Better main loop for this stuff to avoid rounding errors

A quick view of an example I found this happening per frame:

Why create the object and rerun that same unpack logic every time on the same source bytes. Should there be a Clone on the Pigo struct that just copies everything to a new struct to be better performing when running over and over?

Tried in 2 systems and getting the JS errors . Logs attached
localhost-1584545562656.log

Hi there! I'm trying to pass the values for the eyes separately using the blinkdet example, but seems like it outputs the values for both the eyes together and not separately (just outputting max_radius)

Is there a way to output the values that I get using -json via a particular variable or just output the values (max_radius) of both the eyes separately?

cParams := pigo.CascadeParams{
	MinSize:     fd.minSize,
	MaxSize:     fd.maxSize,
	ShiftFactor: fd.shiftFactor,
	ScaleFactor: fd.scaleFactor,
	
	ImageParams: pigo.ImageParams{
		Pixels: pixels,
		Rows:   rows,
		Cols:   cols,
		Dim:    cols,
	},
}

It never appears in README.md apart from these four lines.

The demo documentation declares that there are supported keys:
s - Show/hide pupils
c - Circle through the detection shape types (rectangle|circle|ellipse)
f - Show/hide facial landmark points (hidden by default)

On my local machine, the webcam streaming works and the fps is shown but the keys when pressed don't toggle the features mapped to them.

go version go1.17.2 linux/amd64
Browser: Firefox & Chrome

Hi,

I'm starting to use the Pigo command, and while it's pretty good at detecting faces, its usage is a bit weird. Before that, I've tried some other applications like face_detection and facedetect.
Both can output the coordinates of the faces directly to the console (which is pretty useful to retrieve these infos from PHP).
But Pigo forces to create a PNG file (which is useless for what I want to do), and writes the JSON data to output.json in the current directory. If two calls are made in parallel, it could be a problem.

Would it be possible to:

• make the -out parameter optional?
• output -json to the console if no file is specified, or to the specified file instead of 'output.json'?

Hi,
Thanks for the great work here.

Is it possible to use pigo to extract faces from a photo and save the extracted faces individually.

The number of faces in this image should be 4 rectangles, so why get -out file is 3 rectangles (-iou 0.1 or -iou 0.2)

-iou 0.1

-iou 0.2

In code I can see that DrawRectangles and Face Coordinates have the same input, but it turns out that the coordinates doesn't match face boxes. Is it a bug? Or I don't understand something?

Is your feature request related to a problem? Please describe.

This library is used in PhotoPrism: photoprism/photoprism#22 (comment)

PhotoPrism allows to define a threshold for the score.

While trying to document the value of that threshold, I wonder what are the possible values.

Describe the solution you'd like

Define the minimal and maximal values for the score.

Describe alternatives you've considered

According to this article:

An image region is considered being face if the detection score average is above a certain threshold (in general around 0.995).

The score goes from 0 to 1?

Additional context

.

Issue

The image attached passes facial detection.

API related bug

The images should not be detected as a face

Expected behaviour

No Face Detected

Other Information

• Go version - go version go1.17.1 darwin/amd64

Observation

This image also passes default android face detection but I belive pigo should be better

I use a stream of images and one of the tasks is to group faces, could you recommend a method for comparing faces.

Hey,

I was trying to run facefinder on this particular image

https://user-images.githubusercontent.com/88735/81735832-f0d8cf00-9495-11ea-909c-2bd1dd65394b.png

and I am unable to detect any faces. Can you help me why it's the case?
Actually I saw this pattern for different faces which are quite close such as couples.

Command I used

./pigo --in couple6.png -out out.png -pl -cf cascade/facefinder

I just took the defaults

Is the feature request related to an existing bug? Please describe.

Describe the solution you'd like

Describe alternatives you've considered

Additional context

First, thanks for this, this is very much appreciated, and the code looks solid.

I have one small request, which I think is needed for this to be used as a library. And that is to replace the log.Fatal (which does an ´os.Exit` I don't think is even possible to catch in a client):

https://github.com/esimov/pigo/blob/master/core/pigo.go#L67

I would suggest just returning an error. This may be "this should never happen errors", and then a panic may be in order.

Can this detect a face from a group of given a face to match with ?

Use case is video conferencing and Identification of a user for authentication

First off, just want to appreciate and thank you for this cool project you're working on.
I was wondering if you could add the feature to compare two faces and describe how likely they are.

I tried running pigo with various values for -angle and -scale but couldn't detect the face. Would appreciate any help!

https://github.com/blackjack/webcam can use this。

I got the face detection working (thanks for your help in #12) but when I specify -json, I see output.json getting created but empty.

./pigo -cf data/facefinder -in in.jpg -out out.png -angle=0.3 -json

Hi Endre (@esimov),

We would like to write a short piece about Pigo for our audience of developers and would like to do a Q&A with you. Article would be published on infoq.com. Last article on wasm was about a desktop game editor ported to WebAssembly. Let me know if that is of any interest to you.

Hello, I got the WASM demo working and it's great. Is there a plan / easy way to enable all facial landmarks? I see some code is commented out. Thanks!

Hi,

If I run your web example than I get this error:
2020/04/24 21:51:38 [ERROR] reading next part multipart: NextPart: EOF
I must admit that I don't meet the requirements for this example. My python is python3 and my opencv is 4.2. OS is Ubuntu 20.04LTS.
I tested as follows

1. Run: go run main.go -cf "../../cascade/facefinder"
2. Open URL: localhost:8081/cam in browser
3. Then this error appears
   Any ideas? Thanks

I was wondering if this library can be used for detecting blinks.

You have a real-time example using python. Since chrome now supports (experimental) shape detection and video camera access: https://paul.kinlan.me/face-detection/ & https://medium.com/@joomiguelcunha/lets-play-with-chrome-s-face-detection-api-ca13017a958f
I was wondering if using your WASM port it was able to detect if a face was blinking.