update code

AILab-CVC · Feb 1, 2024 · dc17c77 · dc17c77
1 parent 1325194
commit dc17c77
Show file tree

Hide file tree

Showing 43 changed files with 5,860 additions and 7 deletions.
diff --git a/LICENSE b/LICENSE
diff --git a/README.md b/README.md
@@ -27,11 +27,20 @@
 
 ## Updates 
 
+YOLO-World is under active development and please stay tuned ☕️!
+
 `[2024-1-31]:` We are excited to launch **YOLO-World**, a cutting-edge real-time open-vocabulary object detector.  
-`[2024-2-1]:` We deploy the YOLO-World demo on [HuggingFace 🤗](https://huggingface.co/spaces/stevengrove/YOLO-World), you can try it now!
+`[2024-2-1]:` We deploy the YOLO-World demo on [HuggingFace 🤗](https://huggingface.co/spaces/stevengrove/YOLO-World), you can try it now!  
+`[2024-2-1]:` We've released the code and weights of YOLO-World now!
 
 ## TODO
 
+- [ ] Complete documents for pre-training YOLO-World.
+- [ ] Deployment toolkits, e.g., ONNX or TensorRT. 
+- [ ] Inference acceleration and scripts for speed evaluation.
+- [ ] Automatic labeling framework for image-text pairs, such as CC3M.
+
+
 ## Highlights
 
 This repo contains the PyTorch implementation, pre-trained weights, and pre-training/fine-tuning code for YOLO-World.
@@ -62,14 +71,14 @@ We've pre-trained YOLO-World-S/M/L from scratch and evaluate on the `LVIS val-1.
 
 ### Zero-shot Inference on LVIS dataset
 
-| model | Pre-train Data | AP | AP<sub>r</sub> | AP<sub>c</sub> | AP<sub>f</sub> | weights | log |
-| :---- | :------------- | :-:| :------------: |:-------------: | :-------: | :---: | :---: |
-| [YOLO-World-S]() | O365+GoldG | 26.2 | 19.1 | 23.6 | 29.8  | - | [coming soon] | [log]() |
-| [YOLO-World-M]() | O365+GoldG | 31.0 | 23.8 | 29.2 | 33.9  | - | [coming soon] | [log]() |
-| [YOLO-World-L]() | O365+GoldG | 35.0 | 27.1 | 32.8 | 38.3 | - | [coming soon]| [log]() |
+| model | Pre-train Data | AP<sup>mini</su> | AP<sub>r</sub> | AP<sub>c</sub> | AP<sub>f</sub> | AP<sup>val</su> | AP<sub>r</sub> | AP<sub>c</sub> | AP<sub>f</sub> | weights | log |
+| :---- | :------------- | :-:| :------------: |:-------------: | :------------: | :-:| :------------: |:-------------: | :------------:  | :---: | :---: |  
+| [YOLO-World-S]() | O365+GoldG | 26.2 | 19.1 | 23.6 | 29.8  | - |  | [log]() |
+| [YOLO-World-M]() | O365+GoldG | 31.0 | 23.8 | 29.2 | 33.9  | - | | [log]() |
+| [YOLO-World-L]() | O365+GoldG | 35.0 | 27.1 | 32.8 | 38.3 | - | | [log]() |
 
 **NOTE:**
-1. The evaluation results are tested on LVIS minival in a zero-shot manner.
+1. The evaluation results are tested on LVIS minival ([fixed AP](https://github.com/achalddave/large-vocab-devil) with 1,000 dets) and LVIS val in a zero-shot manner.
 
 
 ### Finetuning on COCO dataset

diff --git a/.../pretrain/yolo_world_l_dual_l2norm_2e-4_100e_4x8gpus_obj365v1_goldg_train_lvis_minival.py b/.../pretrain/yolo_world_l_dual_l2norm_2e-4_100e_4x8gpus_obj365v1_goldg_train_lvis_minival.py
@@ -0,0 +1,173 @@
+_base_ = ('../../third_party/mmyolo/configs/yolov8/'
+          'yolov8_l_syncbn_fast_8xb16-500e_coco.py')
+custom_imports = dict(imports=['yolo_world'],
+                      allow_failed_imports=False)
+
+# hyper-parameters
+num_classes = 1203
+num_training_classes = 80
+max_epochs = 100  # Maximum training epochs
+close_mosaic_epochs = 2
+save_epoch_intervals = 2
+text_channels = 512
+neck_embed_channels = [128, 256, _base_.last_stage_out_channels // 2]
+neck_num_heads = [4, 8, _base_.last_stage_out_channels // 2 // 32]
+base_lr = 2e-3
+weight_decay = 0.05 / 2
+train_batch_size_per_gpu = 16
+
+# model settings
+model = dict(
+    type='YOLOWorldDetector',
+    mm_neck=True,
+    num_train_classes=num_training_classes,
+    num_test_classes=num_classes,
+    data_preprocessor=dict(type='YOLOWDetDataPreprocessor'),
+    backbone=dict(
+        _delete_=True,
+        type='MultiModalYOLOBackbone',
+        image_model={{_base_.model.backbone}},
+        text_model=dict(
+            type='HuggingCLIPLanguageBackbone',
+            model_name='pretrained_models/clip-vit-base-patch32-projection',
+            frozen_modules=['all'])),
+    neck=dict(type='YOLOWolrdDualPAFPN',
+              guide_channels=text_channels,
+              embed_channels=neck_embed_channels,
+              num_heads=neck_num_heads,
+              block_cfg=dict(type='MaxSigmoidCSPLayerWithTwoConv'),
+              text_enhancder=dict(type='ImagePoolingAttentionModule',
+                                  embed_channels=256,
+                                  num_heads=8)),
+    bbox_head=dict(type='YOLOWorldHead',
+                   head_module=dict(type='YOLOWorldHeadModule',
+                                    embed_dims=text_channels,
+                                    num_classes=num_training_classes)),
+    train_cfg=dict(assigner=dict(num_classes=num_training_classes)))
+
+# dataset settings
+text_transform = [
+    dict(type='RandomLoadText',
+         num_neg_samples=(num_classes, num_classes),
+         max_num_samples=num_training_classes,
+         padding_to_max=True,
+         padding_value=''),
+    dict(type='mmdet.PackDetInputs',
+         meta_keys=('img_id', 'img_path', 'ori_shape', 'img_shape', 'flip',
+                    'flip_direction', 'texts'))
+]
+train_pipeline = [
+    *_base_.pre_transform,
+    dict(type='MultiModalMosaic',
+         img_scale=_base_.img_scale,
+         pad_val=114.0,
+         pre_transform=_base_.pre_transform),
+    dict(
+        type='YOLOv5RandomAffine',
+        max_rotate_degree=0.0,
+        max_shear_degree=0.0,
+        scaling_ratio_range=(1 - _base_.affine_scale, 1 + _base_.affine_scale),
+        max_aspect_ratio=_base_.max_aspect_ratio,
+        border=(-_base_.img_scale[0] // 2, -_base_.img_scale[1] // 2),
+        border_val=(114, 114, 114)),
+    *_base_.last_transform[:-1],
+    *text_transform,
+]
+train_pipeline_stage2 = [*_base_.train_pipeline_stage2[:-1], *text_transform]
+obj365v1_train_dataset = dict(
+    type='MultiModalDataset',
+    dataset=dict(
+        type='YOLOv5Objects365V1Dataset',
+        data_root='data/objects365v1/',
+        ann_file='annotations/objects365_train.json',
+        data_prefix=dict(img='train/'),
+        filter_cfg=dict(filter_empty_gt=False, min_size=32)),
+    class_text_path='data/captions/obj365v1_class_captions.json',
+    pipeline=train_pipeline)
+
+mg_train_dataset = dict(type='YOLOv5MixedGroundingDataset',
+                        data_root='data/mixed_grounding/',
+                        ann_file='annotations/final_mixed_train_no_coco.json',
+                        data_prefix=dict(img='gqa/images/'),
+                        filter_cfg=dict(filter_empty_gt=False, min_size=32),
+                        pipeline=train_pipeline)
+
+flickr_train_dataset = dict(
+    type='YOLOv5MixedGroundingDataset',
+    data_root='data/flickr/',
+    ann_file='annotations/final_flickr_separateGT_train.json',
+    data_prefix=dict(img='full_images/'),
+    filter_cfg=dict(filter_empty_gt=True, min_size=32),
+    pipeline=train_pipeline)
+
+train_dataloader = dict(batch_size=train_batch_size_per_gpu,
+                        collate_fn=dict(type='yolow_collate'),
+                        dataset=dict(_delete_=True,
+                                     type='ConcatDataset',
+                                     datasets=[
+                                         obj365v1_train_dataset,
+                                         flickr_train_dataset, mg_train_dataset
+                                     ],
+                                     ignore_keys=['classes', 'palette']))
+
+test_pipeline = [
+    *_base_.test_pipeline[:-1],
+    dict(type='LoadText'),
+    dict(type='mmdet.PackDetInputs',
+         meta_keys=('img_id', 'img_path', 'ori_shape', 'img_shape',
+                    'scale_factor', 'pad_param', 'texts'))
+]
+coco_val_dataset = dict(
+    _delete_=True,
+    type='MultiModalDataset',
+    dataset=dict(type='YOLOv5LVISV1Dataset',
+                 data_root='data/coco/',
+                 test_mode=True,
+                 ann_file='lvis/lvis_v1_minival_inserted_image_name.json',
+                 data_prefix=dict(img=''),
+                 batch_shapes_cfg=None),
+    class_text_path='data/captions/lvis_v1_class_captions.json',
+    pipeline=test_pipeline)
+val_dataloader = dict(dataset=coco_val_dataset)
+test_dataloader = val_dataloader
+
+val_evaluator = dict(type='mmdet.LVISMetric',
+                     ann_file='data/coco/lvis/\
+                         lvis_v1_minival_inserted_image_name.json',
+                     metric='bbox')
+test_evaluator = val_evaluator
+
+# training settings
+default_hooks = dict(param_scheduler=dict(max_epochs=max_epochs),
+                     checkpoint=dict(interval=save_epoch_intervals,
+                                     rule='greater'))
+custom_hooks = [
+    dict(type='EMAHook',
+         ema_type='ExpMomentumEMA',
+         momentum=0.0001,
+         update_buffers=True,
+         strict_load=False,
+         priority=49),
+    dict(type='mmdet.PipelineSwitchHook',
+         switch_epoch=max_epochs - close_mosaic_epochs,
+         switch_pipeline=train_pipeline_stage2)
+]
+train_cfg = dict(max_epochs=max_epochs,
+                 val_interval=10,
+                 dynamic_intervals=[((max_epochs - close_mosaic_epochs),
+                                     _base_.val_interval_stage2)])
+optim_wrapper = dict(optimizer=dict(
+    _delete_=True,
+    type='AdamW',
+    lr=base_lr,
+    weight_decay=weight_decay,
+    batch_size_per_gpu=train_batch_size_per_gpu),
+                     paramwise_cfg=dict(bias_decay_mult=0.0,
+                                        norm_decay_mult=0.0,
+                                        custom_keys={
+                                            'backbone.text_model':
+                                            dict(lr_mult=0.01),
+                                            'logit_scale':
+                                            dict(weight_decay=0.0)
+                                        }),
+                     constructor='YOLOWv5OptimizerConstructor')
diff --git a/.../pretrain/yolo_world_m_dual_l2norm_2e-4_100e_4x8gpus_obj365v1_goldg_train_lvis_minival.py b/.../pretrain/yolo_world_m_dual_l2norm_2e-4_100e_4x8gpus_obj365v1_goldg_train_lvis_minival.py
@@ -0,0 +1,173 @@
+_base_ = ('../../third_party/mmyolo/configs/yolov8/'
+          'yolov8_m_syncbn_fast_8xb16-500e_coco.py')
+custom_imports = dict(imports=['yolo_world'],
+                      allow_failed_imports=False)
+
+# hyper-parameters
+num_classes = 1203
+num_training_classes = 80
+max_epochs = 100  # Maximum training epochs
+close_mosaic_epochs = 2
+save_epoch_intervals = 2
+text_channels = 512
+neck_embed_channels = [128, 256, _base_.last_stage_out_channels // 2]
+neck_num_heads = [4, 8, _base_.last_stage_out_channels // 2 // 32]
+base_lr = 2e-3
+weight_decay = 0.05 / 2
+train_batch_size_per_gpu = 16
+
+# model settings
+model = dict(
+    type='YOLOWorldDetector',
+    mm_neck=True,
+    num_train_classes=num_training_classes,
+    num_test_classes=num_classes,
+    data_preprocessor=dict(type='YOLOWDetDataPreprocessor'),
+    backbone=dict(
+        _delete_=True,
+        type='MultiModalYOLOBackbone',
+        image_model={{_base_.model.backbone}},
+        text_model=dict(
+            type='HuggingCLIPLanguageBackbone',
+            model_name='pretrained_models/clip-vit-base-patch32-projection',
+            frozen_modules=['all'])),
+    neck=dict(type='YOLOWolrdDualPAFPN',
+              guide_channels=text_channels,
+              embed_channels=neck_embed_channels,
+              num_heads=neck_num_heads,
+              block_cfg=dict(type='MaxSigmoidCSPLayerWithTwoConv'),
+              text_enhancder=dict(type='ImagePoolingAttentionModule',
+                                  embed_channels=256,
+                                  num_heads=8)),
+    bbox_head=dict(type='YOLOWorldHead',
+                   head_module=dict(type='YOLOWorldHeadModule',
+                                    embed_dims=text_channels,
+                                    num_classes=num_training_classes)),
+    train_cfg=dict(assigner=dict(num_classes=num_training_classes)))
+
+# dataset settings
+text_transform = [
+    dict(type='RandomLoadText',
+         num_neg_samples=(num_classes, num_classes),
+         max_num_samples=num_training_classes,
+         padding_to_max=True,
+         padding_value=''),
+    dict(type='mmdet.PackDetInputs',
+         meta_keys=('img_id', 'img_path', 'ori_shape', 'img_shape', 'flip',
+                    'flip_direction', 'texts'))
+]
+train_pipeline = [
+    *_base_.pre_transform,
+    dict(type='MultiModalMosaic',
+         img_scale=_base_.img_scale,
+         pad_val=114.0,
+         pre_transform=_base_.pre_transform),
+    dict(
+        type='YOLOv5RandomAffine',
+        max_rotate_degree=0.0,
+        max_shear_degree=0.0,
+        scaling_ratio_range=(1 - _base_.affine_scale, 1 + _base_.affine_scale),
+        max_aspect_ratio=_base_.max_aspect_ratio,
+        border=(-_base_.img_scale[0] // 2, -_base_.img_scale[1] // 2),
+        border_val=(114, 114, 114)),
+    *_base_.last_transform[:-1],
+    *text_transform,
+]
+train_pipeline_stage2 = [*_base_.train_pipeline_stage2[:-1], *text_transform]
+obj365v1_train_dataset = dict(
+    type='MultiModalDataset',
+    dataset=dict(
+        type='YOLOv5Objects365V1Dataset',
+        data_root='data/objects365v1/',
+        ann_file='annotations/objects365_train.json',
+        data_prefix=dict(img='train/'),
+        filter_cfg=dict(filter_empty_gt=False, min_size=32)),
+    class_text_path='data/captions/obj365v1_class_captions.json',
+    pipeline=train_pipeline)
+
+mg_train_dataset = dict(type='YOLOv5MixedGroundingDataset',
+                        data_root='data/mixed_grounding/',
+                        ann_file='annotations/final_mixed_train_no_coco.json',
+                        data_prefix=dict(img='gqa/images/'),
+                        filter_cfg=dict(filter_empty_gt=False, min_size=32),
+                        pipeline=train_pipeline)
+
+flickr_train_dataset = dict(
+    type='YOLOv5MixedGroundingDataset',
+    data_root='data/flickr/',
+    ann_file='annotations/final_flickr_separateGT_train.json',
+    data_prefix=dict(img='full_images/'),
+    filter_cfg=dict(filter_empty_gt=True, min_size=32),
+    pipeline=train_pipeline)
+
+train_dataloader = dict(batch_size=train_batch_size_per_gpu,
+                        collate_fn=dict(type='yolow_collate'),
+                        dataset=dict(_delete_=True,
+                                     type='ConcatDataset',
+                                     datasets=[
+                                         obj365v1_train_dataset,
+                                         flickr_train_dataset, mg_train_dataset
+                                     ],
+                                     ignore_keys=['classes', 'palette']))
+
+test_pipeline = [
+    *_base_.test_pipeline[:-1],
+    dict(type='LoadText'),
+    dict(type='mmdet.PackDetInputs',
+         meta_keys=('img_id', 'img_path', 'ori_shape', 'img_shape',
+                    'scale_factor', 'pad_param', 'texts'))
+]
+coco_val_dataset = dict(
+    _delete_=True,
+    type='MultiModalDataset',
+    dataset=dict(type='YOLOv5LVISV1Dataset',
+                 data_root='data/coco/',
+                 test_mode=True,
+                 ann_file='lvis/lvis_v1_minival_inserted_image_name.json',
+                 data_prefix=dict(img=''),
+                 batch_shapes_cfg=None),
+    class_text_path='data/captions/lvis_v1_class_captions.json',
+    pipeline=test_pipeline)
+val_dataloader = dict(dataset=coco_val_dataset)
+test_dataloader = val_dataloader
+
+val_evaluator = dict(type='mmdet.LVISMetric',
+                     ann_file='data/coco/lvis/\
+                         lvis_v1_minival_inserted_image_name.json',
+                     metric='bbox')
+test_evaluator = val_evaluator
+
+# training settings
+default_hooks = dict(param_scheduler=dict(max_epochs=max_epochs),
+                     checkpoint=dict(interval=save_epoch_intervals,
+                                     rule='greater'))
+custom_hooks = [
+    dict(type='EMAHook',
+         ema_type='ExpMomentumEMA',
+         momentum=0.0001,
+         update_buffers=True,
+         strict_load=False,
+         priority=49),
+    dict(type='mmdet.PipelineSwitchHook',
+         switch_epoch=max_epochs - close_mosaic_epochs,
+         switch_pipeline=train_pipeline_stage2)
+]
+train_cfg = dict(max_epochs=max_epochs,
+                 val_interval=10,
+                 dynamic_intervals=[((max_epochs - close_mosaic_epochs),
+                                     _base_.val_interval_stage2)])
+optim_wrapper = dict(optimizer=dict(
+    _delete_=True,
+    type='AdamW',
+    lr=base_lr,
+    weight_decay=weight_decay,
+    batch_size_per_gpu=train_batch_size_per_gpu),
+                     paramwise_cfg=dict(bias_decay_mult=0.0,
+                                        norm_decay_mult=0.0,
+                                        custom_keys={
+                                            'backbone.text_model':
+                                            dict(lr_mult=0.01),
+                                            'logit_scale':
+                                            dict(weight_decay=0.0)
+                                        }),
+                     constructor='YOLOWv5OptimizerConstructor')