Thanks for the great work.

Would you be willing to add trained weights (and possibly a single-file out-of-sample classification demo) to the repo? This way people could directly apply your best classifier to a photo of their choice without having to retrain the whole model themselves.

I dont know about annotations_epi.json
what and where is annotations_epi.json file?
Thank you very much

I wish to use TACO Bounding boxes as well for my project. Would you mind sharing the converted annotations?

To run train.py of EfficientDet the path_to_all images means that I have to take the images of each dataset and join in just one folder ?

python3 train.py path_to_all_images
--ann_name ../annotations/binary_mixed --model tf_efficientdet_d2
--batch-size 4 --decay-rate 0.95 --lr .001 --workers 4 --warmup-epochs 5
--model-ema --dataset multi --pretrained --num-classes 1 --color-jitter 0.1
--reprob 0.2 --epochs 20 --device cuda:0

Thanks for your great projects and Thanks for share the code of your great work!
I run the openlittermap_downloader.py.but I can't download the database.
can you give me some advice?thanks very much!!!

Hi, thanks for sharing this amazing work. I could not find the trained model weights anywhere. Have you provided them or do you plan to share them?

Hello, thank you for the great work. I am new learner of python, I am doing a bootcamp and want to use your project as demo. I installed the related packages and want to run the file make_predictions.py. However, I am not sure what should I provide in the parameter --detector.

For the another similar parameter --classifier, I downloaded the .pth.tar file which you provided in Google Drive. I am not sure if it is correct as well.

Thanks for such a great work again. I learn a lot through your detailed documentation and It would be great if you can provide some helps to me. =)

Hello. I am building an object detector based on the YOLO family of models, and i would like to know how to use this dataset.
Specifically the TACO box dataset annotations, but I'm confused on how to use it. Could you help me? Thanks!

Hi, your project detect-waste requires "funcy==1.15" in its dependency. After analyzing the source code, we found that some other versions of funcy can also be suitable without affecting your project, i.e., funcy 1.16, 1.17. Therefore, we suggest to loosen the dependency on funcy from "funcy==1.15" to "funcy>=1.15,<=1.17" to avoid any possible conflict for importing more packages or for downstream projects that may use detect-waste.

May I pull a request to loosen the dependency on funcy?

By the way, could you please tell us whether such dependency analysis may be potentially helpful for maintaining dependencies easier during your development?

For your reference, here are details in our analysis.

Your project detect-waste(commit id: 6113385) directly uses 3 APIs from package funcy.

funcy.seqs.lfilter, funcy.seqs.lremove, funcy.seqs.lmap

From which, 14 functions are then indirectly called, including 9 funcy's internal APIs and 5 outsider APIs, as follows (neglecting some repeated function occurrences).

[/wimlds-trojmiasto/detect-waste]
+--funcy.seqs.lfilter
|      +--funcy.compat.lfilter
|      |      +--itertools.ifilter
|      +--funcy.funcmakers.make_pred
|      |      +--funcy.funcmakers.make_func
|      |      |      +--funcy.strings.re_tester
|      |      |      |      +--re.compile
|      |      |      +--funcy.strings.re_finder
|      |      |      |      +--funcy.strings._prepare
|      |      |      |      |      +--re.compile
|      |      |      |      |      +--funcy.strings._make_getter
|      |      |      |      |      |      +--operator.methodcaller
|      |      |      +--operator.itemgetter
+--funcy.seqs.lremove
|      +--funcy.seqs.remove
|      |      +--funcy.funcmakers.make_pred
+--funcy.seqs.lmap
|      +--funcy.compat.lmap
|      |      +--itertools.imap
|      +--funcy.funcmakers.make_func

We scan funcy's versions among [1.16, 1.17] and 1.15, the changing functions (diffs being listed below) have none intersection with any function or API we mentioned above (either directly or indirectly called by this project).

diff: 1.15(original) 1.16
['funcy.colls.has_path', 'funcy.calc.CacheMemory.__setitem__', 'funcy.decorators.has_1pos_and_kwonly', 'funcy.calc.CacheMemory.expire', 'funcy.calc.memoize', 'funcy.calc.SkipMemoization', 'funcy.flow.throttle', 'funcy.calc.CacheMemory', 'funcy._inspect.get_spec', 'funcy.decorators.decorator', 'funcy.calc.CacheMemory.clear', 'funcy.flow._ensure_exceptable', 'funcy.flow.reraise', 'funcy.funcs.curry', 'funcy.decorators.has_single_arg', 'funcy.calc.CacheMemory.__init__', 'funcy.calc._memory_decorator', 'funcy.funcs.autocurry', 'funcy.calc.SkipMemory', 'funcy.calc.CacheMemory.__getitem__', 'funcy.calc.cache', 'funcy.flow._is_exception_type', 'funcy.funcs.rcurry']

diff: 1.15(original) 1.17
['funcy.colls.has_path', 'funcy.calc.CacheMemory.__setitem__', 'funcy.colls.zip_values', 'funcy.decorators.Call.__str__', 'funcy.colls.zip_dicts', 'funcy.decorators.has_1pos_and_kwonly', 'funcy.calc.CacheMemory.expire', 'funcy.calc.memoize', 'funcy.calc.SkipMemoization', 'funcy.flow.throttle', 'funcy.calc.CacheMemory', 'funcy._inspect.get_spec', 'funcy._inspect._sig_to_spec', 'funcy.decorators.decorator', 'funcy.calc.CacheMemory.clear', 'funcy.flow._ensure_exceptable', 'funcy.flow.reraise', 'funcy.funcs.curry', 'funcy.decorators.has_single_arg', 'funcy.calc.CacheMemory.__init__', 'funcy.flow.limit_error_rate', 'funcy.calc._memory_decorator', 'funcy.funcs.autocurry', 'funcy.decorators.Call', 'funcy.calc.SkipMemory', 'funcy.calc.CacheMemory.__getitem__', 'funcy.colls.del_in', 'funcy.decorators.Call.__repr__', 'funcy.calc.cache', 'funcy.flow._is_exception_type', 'funcy.funcs.rcurry']

As for other packages, the APIs of @outside_package_name are called by funcy in the call graph and the dependencies on these packages also stay the same in our suggested versions, thus avoiding any outside conflict.

Therefore, we believe that it is quite safe to loose your dependency on funcy from "funcy==1.15" to "funcy>=1.15,<=1.17". This will improve the applicability of detect-waste and reduce the possibility of any further dependency conflict with other projects/packages.

Hi,
Thanks for your wonderful work in the domain. I have worked in underwater waste detection and now i want to use your dataset(waste-classify and waste detect)for my model. Can you guide me how and where to get these datasets.

Thanks again

I noted that https://detectwaste.ml/ is offline. The netlify app is still online https://detectwaste.netlify.app/. Did the domain name expire?

Do you have plan to provide the additional data collected which downloads from google image download?
In addition, what keywords do you use for searching in google image download?

Hi @AgaMiko, It would be really kind if you could help me through steps mentioned in the readme file, under Data Processing - Multiclass training; when I try to run python3 annotations_preprocessing.py I am facing this error:
FileNotFoundError: [Errno 2] No such file or directory: '/dih4/dih4_2/wimlds/TACO-master/data/annotations.json'

Dear wimlds,
First of all, thanks for the great works,

I trained EfficientNet model for waste detection with "tf_efficientdet_d2" (multiclass training)

But when i evaluate the model with demo.py, An error occurred during the model loading.

The error is as below,

$> python3 demo.py --save ./results/image.png --checkpoint /home/jupyter/detect-waste/efficientdet/output/train/20220415-025926-tf_efficientdet_d2/checkpoint-0.pth.tar

Traceback (most recent call last):
File "demo.py", line 268, in
main(args)
File "demo.py", line 235, in main
model = set_model("tf_efficientdet_d2", num_classes, args.checkpoint, args.device)
File "demo.py", line 221, in set_model
checkpoint_path=checkpoint_path
File "/home/jupyter/detect-waste/efficientdet/effdet/factory.py", line 14, in create_model
checkpoint_path=checkpoint_path, checkpoint_ema=checkpoint_ema, **kwargs)
File "/home/jupyter/detect-waste/efficientdet/effdet/factory.py", line 47, in create_model_from_config
load_checkpoint(model, checkpoint_path, use_ema=checkpoint_ema)
File "/opt/conda/lib/python3.7/site-packages/timm/models/helpers.py", line 64, in load_checkpoint
model.load_state_dict(state_dict, strict=strict)
File "/opt/conda/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1498, in load_state_dict
self.class.name, "\n\t".join(error_msgs)))
RuntimeError: Error(s) in loading state_dict for EfficientDet:
size mismatch for class_net.predict.conv_pw.weight: copying a param with shape torch.Size([63, 112, 1, 1]) from checkpoint, the shape in current model is torch.Size([9, 112, 1, 1]).
size mismatch for class_net.predict.conv_pw.bias: copying a param with shape torch.Size([63]) from checkpoint, the shape in current model is torch.Size([9]).

Can you give some advices to solve this error.

Thanks.
Jong Wuk Son

Dear wimlds,
Thanks for your greate effort

I try the source code, but how to create classify-waste dataset.
I cant preproduce it now. Can you share me your classify-waste dataset for training

Thanks,
Xuan Tran

can any one post an video tutorial about this project how to install and run

Hi,

It seem pre weight file was automatically removed from https://drive.google.com/drive/u/0/folders/1wNWoH8rdkG05sBw-OCXp3J73uJPxhcxH
Really appreciate your effort to get it back

Regards
Quang