Commit Graph

Select branches

Hide Pull Requests

datastructure

master

Mono Color

• 573665d3f1 hint added where rise training dataset is located master Till Raab 2024-01-19 11:24:23 +0100
• 8cd422461a detection of rise time based on surpass of tau freq or steepest slope Till Raab 2023-12-04 12:22:16 +0100
• cc4408d75a inference plots reworked Till Raab 2023-12-04 11:25:56 +0100
• 22b05aec76 corrected labels that are checked are now exported to the train dataset via copy instead of moving them. Till Raab 2023-12-04 08:52:26 +0100
• 59ffd101f1 when labels are gererated for an already existing dataset delete the file_dict.csv file used by corret_bboxes Till Raab 2023-12-04 08:50:53 +0100
• 5c74ac463b assignement of bboxes to tracks complete. Correct assignement of chirpt start still necessary... increase to >37% of f1-f0 range suggested Till Raab 2023-11-30 15:19:33 +0100
• 866fd3081d elimination of double detection works Till Raab 2023-11-29 15:00:02 +0100
• f70e74f5e1 inference.py now also writes bbox scores; extract_from_bbox.py detects overlapping bbox groups and eliminates them based on score untill all bboxes have overlap below th... detection of groups is done using time as parameter only. implement also frequency. then sort them as stated previously Till Raab 2023-11-28 14:53:37 +0100
• 93269a96a1 transfere bboxes to time-frequency points Till Raab 2023-11-21 15:20:44 +0100
• 0f0128439e bf Till Raab 2023-11-16 13:30:26 +0100
• 324c690841 model and datasets adapted to new yolo format. reinforced learning is now implemented. correct_bboxes.py now capable of exporting data to train file Till Raab 2023-11-16 12:19:57 +0100
• 3ed6f1102c dont for today. gui is more advanced. work on import/export Till Raab 2023-11-15 17:32:43 +0100
• 3f235f02f3 file dict ssaved and loaded Till Raab 2023-11-15 09:22:41 +0100
• 33f6bb8683 working on bbox correct gui Till Raab 2023-11-14 12:22:48 +0100
• f286156607 work im progress ... Till Raab 2023-11-13 15:10:37 +0100
• 397be5c51b bbox correcter Till Raab 2023-11-08 16:04:04 +0100
• c08ede1e76 working on bbox corrector Till Raab 2023-11-08 15:31:17 +0100
• 5f270c41f2 labels in generate dataset should be int now datastructure Till Raab 2023-11-08 14:04:27 +0100
• 92ab342a65 added miussing dataset generator Till Raab 2023-11-07 16:00:14 +0100
• 9dcf54c139 adapted data structure Till Raab 2023-11-07 15:58:10 +0100
• 28554efe2b reworking the data structure. dataset generator is now adapted... dataloader fn are next Till Raab 2023-11-06 14:19:09 +0100
• 61d5a7246b fixed the validation output. generation of folder was missing Till Raab 2023-11-06 10:57:04 +0100
• 4e1784a399 getting back. start to implement greyscale images as inputs Till Raab 2023-11-06 08:43:25 +0100
• e23301c0d1 README update Till Raab 2023-10-27 13:12:24 +0200
• 3dbc78685d small documentation progress Till Raab 2023-10-27 13:04:53 +0200
• e2b540c329 suggestions for after my hollidays Till Raab 2023-10-27 12:52:37 +0200
• a3382182e5 validation output pics now in validationaoutput/<dataset_name> Till Raab 2023-10-27 11:59:01 +0200
• 0befddab8d bf Till Raab 2023-10-27 11:32:48 +0200
• 19a7266c17 embed for bf Till Raab 2023-10-27 11:31:10 +0200
• 95879e4759 create dataset path in outdirs if not existing Till Raab 2023-10-27 11:30:09 +0200
• 323d79b27a bf Till Raab 2023-10-27 11:27:31 +0200
• f3e88078c4 bf in InferenceDataset in datasets.py Till Raab 2023-10-27 11:25:32 +0200
• 873758d461 embed for bf Till Raab 2023-10-27 11:22:59 +0200
• 71030bb707 embed for bf Till Raab 2023-10-27 11:22:00 +0200
• e3c0286d21 bf Till Raab 2023-10-27 11:20:38 +0200
• cbbf839e72 bf Till Raab 2023-10-27 11:16:51 +0200
• 3690907f93 inference rework Till Raab 2023-10-27 11:15:53 +0200
• 43f0e0a27d dataset files are now more consistend in their names ... more digits for the time stuff Till Raab 2023-10-27 10:28:09 +0200
• 2a0a1a2766 gnerate dataset now capable of only producing images Till Raab 2023-10-27 10:26:40 +0200
• a6641accaf infered img not with suffix: _predicted.png Till Raab 2023-10-27 09:35:29 +0200
• 1042e44f37 inference.py will be rewritten to really infere images without csv files Till Raab 2023-10-27 09:33:47 +0200
• cc6e97c2c8 inference code copied here ... this accually was the validation step with plot output. Till Raab 2023-10-27 09:33:07 +0200
• 0243f560be TRAIN_DIR now named DATA_DIR Till Raab 2023-10-26 13:10:25 +0200
• a91e61f514 stupid bug fixed Till Raab 2023-10-26 12:51:45 +0200
• 5898a885fd bf Till Raab 2023-10-26 12:50:50 +0200
• 030dcfab43 changed TRAIN_DIR to DATA_DIR in all files since it is more representative in this code. test and train datasets are defined by csv files Till Raab 2023-10-26 12:47:29 +0200
• 16bca48cbe README update Till Raab 2023-10-26 12:09:16 +0200
• 214a4bd10c config rework Till Raab 2023-10-26 12:04:35 +0200
• 1258bcf1bd bf Till Raab 2023-10-26 12:01:40 +0200
• 5c2fc32cea README with ToDos implemented Till Raab 2023-10-26 11:59:56 +0200
• 4bbf3607e9 small fix Till Raab 2023-10-26 11:56:14 +0200
• c2256c7001 README initialized Till Raab 2023-10-26 11:54:44 +0200
• 6779203dc4 bf Till Raab 2023-10-26 10:20:19 +0200
• 593e8860d1 embed for figsave bf Till Raab 2023-10-26 10:14:44 +0200
• 76cee2c050 inference images are now saved in a corresponding folder Till Raab 2023-10-26 10:13:48 +0200
• 391379551a check image save process Till Raab 2023-10-26 10:06:55 +0200
• 7f2d1cfb33 fixed image name problematic Till Raab 2023-10-26 09:35:15 +0200
• 0adfc22ec4 check if i can leave image name in target as return from dataset ... to(DEVICE) problematic Till Raab 2023-10-26 09:25:02 +0200
• 72c58af7e4 preicted boxes in dashed green Till Raab 2023-10-26 09:05:08 +0200
• 4373b7c453 bug fixed ... no double images any longer... repeat trainging because there we have the same issue... Till Raab 2023-10-26 09:00:38 +0200
• 9905b6039e big bug... not unique imaged :D Till Raab 2023-10-26 08:56:43 +0200
• efd860a127 bf Till Raab 2023-10-26 08:49:40 +0200
• 10f8f872f6 bf Till Raab 2023-10-26 08:49:00 +0200
• 47cd574354 bf Till Raab 2023-10-26 08:47:30 +0200
• 68b3967a21 added embed for bugfix Till Raab 2023-10-26 08:41:57 +0200
• 7386bd83a5 bf Till Raab 2023-10-26 08:40:44 +0200
• b89cf321f9 bf Till Raab 2023-10-26 08:40:24 +0200
• b1e5b9e665 inference updated Till Raab 2023-10-26 08:38:14 +0200
• fe6e51d625 inference function with dataloader to optimizer inference. now generate plots ... Till Raab 2023-10-26 07:35:41 +0200
• f507877624 bf Till Raab 2023-10-26 07:30:50 +0200
• a98f488813 check if model still finds stuff Till Raab 2023-10-26 07:30:19 +0200
• e1a97ac493 train test split implemented using different csv files Till Raab 2023-10-25 15:31:00 +0200
• ecf110e051 working on the dataset. now i will rework the dataloader to load test and train data based on the corresponding csv files Till Raab 2023-10-25 15:04:09 +0200
• ad74322f94 the model is working. however, in need to generate it and split data in test and train beforehand ... Till Raab 2023-10-25 09:09:20 +0200
• 8a6e7df57f generated huge dataset Till Raab 2023-10-24 15:15:19 +0200
• 48d84d3793 looks better ... write inference code and feed the model Till Raab 2023-10-24 09:17:31 +0200
• 85e675fb48 something works !!! Till Raab 2023-10-24 08:56:35 +0200
• 30a9e71e76 training loop runs ... no feed it Till Raab 2023-10-23 15:26:39 +0200
• 3f66b4a39a alot of work. dataset and loader not created Till Raab 2023-10-23 14:43:01 +0200
• 5ca527e39f generate DataFrame and collect pic names and bounding boxes for training dataet Till Raab 2023-10-20 13:30:27 +0200
• e139a7a283 bf Till Raab 2023-10-20 13:27:45 +0200
• 84b104c6ec bf Till Raab 2023-10-20 13:24:39 +0200
• ebf9690ad3 zorder of bboxes changed by adding them all together in a patch collection Till Raab 2023-10-20 13:22:55 +0200
• 1bb1de442f bug search Till Raab 2023-10-20 13:16:50 +0200
• a31904dd7c bug search Till Raab 2023-10-20 13:14:57 +0200
• c91760db78 bug search Till Raab 2023-10-20 13:13:57 +0200
• e70fce7f98 bf Till Raab 2023-10-20 13:12:05 +0200
• 7d03351057 bf Till Raab 2023-10-20 13:11:19 +0200
• f828977837 bf Till Raab 2023-10-20 13:10:06 +0200
• eafbaef4c6 bboxes: first implementation Till Raab 2023-10-20 13:09:08 +0200
• 29f693554e bf Till Raab 2023-10-20 12:58:44 +0200
• cd7cce01ba bf Till Raab 2023-10-20 12:55:50 +0200
• f08bca9387 bf Till Raab 2023-10-20 12:54:56 +0200
• 59719dd440 embed for bbox exploration Till Raab 2023-10-20 12:53:53 +0200
• 3e6f25820e embed for bbox exploration Till Raab 2023-10-20 12:31:38 +0200
• f051e07035 bg Till Raab 2023-10-20 11:35:20 +0200
• f1aaf4fb37 bf Till Raab 2023-10-20 11:32:31 +0200
• 68d810e693 bf and automatic plot show for 2 sec and then close and contiunue Till Raab 2023-10-20 11:31:38 +0200
• e1ce6c4d80 added rise size >= 10 as condition Till Raab 2023-10-20 11:29:41 +0200
• 9bdfda5543 bf Till Raab 2023-10-20 11:26:44 +0200