Low Light Image Enhancement using Python & Deep Learning

In this video, I'll show you how you can enhance low-light images using TensorFlow & TensorFlow Lite using MIRNet model as proposed by Learning Enriched Features for Real Image Restoration and Enhancement. This model is capable of enhancing low-light images up to a great extent.

Join this channel to get access to perks:

If you do have any questions with what we covered in this video then feel free to ask in the comment section below & I'll do my best to answer those.

If you enjoy these tutorials & would like to support them then the easiest way is to simply like the video & give it a thumbs up & also it's a huge help to share these videos with anyone who you think would find them useful.

Please consider clicking the SUBSCRIBE button to be notified for future videos & thank you all for watching.

You can find me on:

#low-light-image #enhancement

[oshiya.r.a.]

Sir, I'm unable to access the google drive link for training set. It says access denied when I tried to run the code given in github.

[prathmeshpatil]

I was looking for something like this..thanks for the video!

[helifynoe]

Darkness can change things. There it was, just across the street from me late at night, a store sign that said in big letters, HOPPERS RUG ART. But at a later date in the day time it now appeared as SHOPPERS DRUG MART. Yes, 3 of the letters were not lighting up during the dark hours.

[mhmahfuz]

showing this - Access denied with the following error:

Cannot retrieve the public link of the file. You may need to change
the permission to 'Anyone with the link', or have had many accesses.

[principalafmr]

Hi I tried to run the model but it is giving the following error

ValueError: Input 0 is incompatible with layer model: expected shape=(None, None, None, 3), found shape=(None, 3572, 2524, 4)

What to do?

[yasminbanu]

Yes sir it's very informative .But sir how it
could be done with multiple images rather passing single image

[RupaB-ugzg]

Please help, i am not unable to run download model section. It give error in public link to see any one

[thamaraim]

Sir, Initially, the code worked well. Later, it shows errors like" name 'inferer' is not defined "and after that, loading of the MIRNET WEIGHTS also not working. It shows some links for weights downloading.But using that link also does not possible load weights. Please reply

[surfingmoose]

the first image i thought was not that good. the second one was quite good. however the last one, you can see cloud details in the original image and all gone after being processed, plus it makes the image look like the foreground was cutout from a photograph. nice tutorial though.

[farhannurhakim]

Sir are there any tutorial for enhancing underwater image?

[Hamouda]

Thank you for this explanation. You use jupyter IDE for coding? Thkx

[sushobhanmudi]

can video enhancement be possible using the same process? and how? please reply

[dynasty]

What kind of latency can we expect in real-time use cases?

[chitti]

sir when we try load the weights, it s giving this message

Access denied with the following error:

Cannot retrieve the public link of the file. You may need to change
the permission to 'Anyone with the link', or have had many accesses.

pls can you help in resolving it sir

[zshan]

Hey, thanks bhavesh. Is there any other models for the video restoration and enhancement.

[ipvikas]

I am getting an error: ValueError: Dimensions must be equal, but are 782 and 780 for '{{node model_1/add_180/add_1}} = AddV2[T=DT_FLOAT](model_1/add_180/add, model_1/add_179/add)' with input shapes: [?, 782, 782, 64], [?, 780, 780, 64].

[athulbs]

Is this the program ofEnhancement of foggy images??

[annisawidiyanti]

Hello sir, thank you for this video it really helpful as a reference, butt sorry can i help you i really need for video enhancement using python. Thankyou

[karmanyabattish]

hi it is showing me error that the jpg file not found what should i do please help

[jriegjg]

Hello, I want help... I want to work on a topic about image enhancement using deep learning. I haven't found simplified educational resources to understand the process from scratch. I would appreciate your assistance if possible, by providing me with good sources. Thank you very much Pr durand