automl
automl copied to clipboard
google
Google Brain AutoML
Hi there, thanks for the nice script provided first. when I applied efficientdet D4 on my own dataset, I met two questions, does anyone can help me with that: 1....
Follow the tutorial 8 Finetune on PASCAL VOC 2012 with detector COCO ckpt, I want to train efficientdet-d0 on my data. When I set var_freeze_expr: '(efficientnet|fpn_cells|resample_p6)' in the configure file,...
Use standard file utilities to get mtimes. ERROR:tensorflow:Model diverged with loss = NaN. E0513 16:02:25.483811 139845663950592 basic_session_run_hooks.py:760] Model diverged with loss = NaN. ERROR:tensorflow:Error recorded from training_loop: NaN loss during...
Hi, I am trying to export a frozen graph which has only a neural network nodes. To do this, I turn on the export mode and set the only_network flag...
Hi, @fsx950223 @mingxingtan Augmenator class will take config and will build augmentation pipeline, Something like **policies in autoaugment** this will require registry file which we can decorate on augmentation class...
QUESTION: Can EfficientDet Perform Multi-Label Instance Segmentation Or Only Semantic Segmentation
As the title states. Can we use the **`keras/segmentation.py`** file to perform instance segmentation (single or multi-label) or just semantic segmentation? While it appears other closed answers/issues reference the ability...
I am following section "5. Inference for images" of README and I tried it with two models: D0 and D7. Unexpectedly the result of D0 is much better than D7....
From my experiments, I found that a small model (EfficientDet-d3) can detect small objects better than the bigger model (EfficientDet-d4, d5), Although the APs score of d4 is better than...
Hi, @mingxingtan @fsx950223 https://arxiv.org/pdf/2012.07177v1.pdf @mingxingtan You must know this, as it's from the google brain team. Hope this can help. p.s: I can help to create the script:) Thanks
Hi, I have one question about match/unmatched_threshold of anchor labeler. Why unmatched_threshold is same value as matched_threshold? retinanet assign 0.5/0.4 to the match/unmatched_threshold. If there is a reason to change...
