Merge Back 2.3.0 (+2.2.0.rc12) to develop

.ci/docker/Dockerfile

@@ -5,16 +5,9 @@
 FROM nvidia/cuda:12.1.0-runtime-ubuntu20.04@sha256:c1869c30f46fff478a37ed58d9dace7e08519541274f03424d0b78bd35b2c73a AS python_base_cuda
 LABEL maintainer="OpenVINO Training Extensions Development Team"
 
-ARG HTTP_PROXY
-ARG HTTPS_PROXY
-ARG NO_PROXY
 ARG uid
 ARG gid
 
-# Setup proxies
-ENV http_proxy=$HTTP_PROXY
-ENV https_proxy=$HTTPS_PROXY
-ENV no_proxy=$NO_PROXY
 ENV DEBIAN_FRONTEND="noninteractive"
 
 # hadolint ignore=DL3008

.ci/docker/build.sh

@@ -53,9 +53,6 @@ fi
 TAG=$1
 
 docker build -f ./Dockerfile \
---build-arg HTTP_PROXY="${http_proxy:?}" \
---build-arg HTTPS_PROXY="${https_proxy:?}" \
---build-arg NO_PROXY="${no_proxy:?}" \
 --build-arg ACTIONS_RUNNER_VER="$ACTIONS_RUNNER_VER" \
 --build-arg gid="$(id -g)" \
 --build-arg uid="$UID" \

.ci/docker/start-runner.sh

@@ -1,7 +1,7 @@
 #!/bin/bash
 
 GPU_ID="all"
-VER_CUDA="11.7.1"
+VER_CUDA="12.1.0"
 TAG_RUNNER="latest"
 ADDITIONAL_LABELS=""
 MOUNT_PATH=""
@@ -149,9 +149,6 @@ if [ "$DEBUG_CONTAINER" = true ]; then
         --name "$CONTAINER_NAME" \
         -e NVIDIA_VISIBLE_DEVICES="$GPU_ID" \
         ${ENV_FLAGS} \
-        -e http_proxy=http://proxy-chain.intel.com:911 \
-        -e https_proxy=http://proxy-chain.intel.com:912 \
-        -e no_proxy=intel.com,.intel.com,localhost,127.0.0.0/8 \
         ${MOUNT_FLAGS} \
         ${CACHE_MOUNT_FLAGS} \
         "$DOCKER_REG_ADDR"/ote/ci/cu"$VER_CUDA"/runner:"$TAG_RUNNER"; RET=$?
@@ -172,9 +169,6 @@ else
         --name "$CONTAINER_NAME" \
         -e NVIDIA_VISIBLE_DEVICES="$GPU_ID" \
         ${ENV_FLAGS} \
-        -e http_proxy=http://proxy-chain.intel.com:911 \
-        -e https_proxy=http://proxy-chain.intel.com:912 \
-        -e no_proxy=intel.com,.intel.com,localhost,127.0.0.0/8 \
         ${MOUNT_FLAGS} \
         ${CACHE_MOUNT_FLAGS} \
         "$DOCKER_REG_ADDR"/ote/ci/cu"$VER_CUDA"/runner:"$TAG_RUNNER"; RET=$?

.github/workflows/pre_merge.yaml

@@ -101,9 +101,6 @@ jobs:
           - task: "multi_cls_classification"
           - task: "multi_label_classification"
           - task: "hlabel_classification"
-          - task: "detection"
-          - task: "instance_segmentation"
-          - task: "semantic_segmentation"
           - task: "visual_prompting"
           - task: "zero_shot_visual_prompting"
           - task: "anomaly_classification"
@@ -127,3 +124,32 @@ jobs:
           rm /tmp/requirements.txt
       - name: Run Integration Test
         run: tox -vv -e integration-test-${{ matrix.task }}
+  Integration-Test-Large:
+    if: |
+      github.event.pull_request.draft == false &&
+      !(startsWith(github.event.pull_request.title, '[WIP]'))
+    runs-on: [self-hosted, linux, x64, dev, dmount]
+    needs: Unit-Test
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - task: "detection"
+          - task: "instance_segmentation"
+          - task: "semantic_segmentation"
+    name: Integration-Test-Large-${{ matrix.task }}-py310
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@b4ffde65f46336ab88eb53be808477a3936bae11 # v4.1.1
+      - name: Install Python
+        uses: actions/setup-python@39cd14951b08e74b54015e9e001cdefcf80e669f # v5.1.1
+        with:
+          python-version: "3.10"
+      - name: Install tox
+        run: |
+          python -m pip install --require-hashes --no-deps -r .ci/requirements.txt
+          pip-compile --generate-hashes --output-file=/tmp/requirements.txt --extra=ci_tox pyproject.toml
+          python -m pip install --require-hashes --no-deps -r /tmp/requirements.txt
+          rm /tmp/requirements.txt
+      - name: Run Integration Test
+        run: tox -vv -e integration-test-${{ matrix.task }}

CHANGELOG.md

@@ -7,76 +7,80 @@ All notable changes to this project will be documented in this file.
 ### New features
 
 - Turn on/off classification augmentations
-  (https://github.com/openvinotoolkit/training_extensions/pull/4039)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4039>)
 
 ### Enhancements
 
 - Update visual prompting pipeline for multi-label zero-shot learning support
-  (https://github.com/openvinotoolkit/training_extensions/pull/3993)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3993>)
 - Update to work torch compile in detection
-  (https://github.com/openvinotoolkit/training_extensions/pull/4003)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4003>)
 - Fix MaskRCNN/RTMDet-Inst/MaskRCNNTV Explain Mode
-  (https://github.com/openvinotoolkit/training_extensions/pull/4053)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4053>)
 
 ## \[2.3.0\]
 
 ### New features
 
 - Add YOLOv9 model for Object Detection
-  (https://github.com/openvinotoolkit/training_extensions/pull/3917)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3917>, <https://github.com/openvinotoolkit/training_extensions/pull/4026>)
 - Add OV inference for keypoint detection
-  (https://github.com/openvinotoolkit/training_extensions/pull/3970)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3970>)
 - Add tiling for semantic segmentation
-  (https://github.com/openvinotoolkit/training_extensions/pull/3954)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3954>)
 - Add 3D Object Detection task with MonoDETR model
-  (https://github.com/openvinotoolkit/training_extensions/pull/3979)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3979>)
+- Add OpenVINO inference for 3D Object Detection task
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4017>)
 
 ### Enhancements
 
 - Upgrade OV, MAPI, and NNCF dependencies
-  (https://github.com/openvinotoolkit/training_extensions/pull/3967)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3967>)
 - Instance Segmentation Model refactoring
-  (https://github.com/openvinotoolkit/training_extensions/pull/3865)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3865>)
 - Bump torch and lightning to 2.4.0 versions
-  (https://github.com/openvinotoolkit/training_extensions/pull/3843)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3843>)
 - Add mAP metric to evaluate multilabel classification
-  (https://github.com/openvinotoolkit/training_extensions/pull/3985)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3985>)
 
 ### Bug fixes
 
 - Fix a wrong HPO log
-  (https://github.com/openvinotoolkit/training_extensions/pull/3972)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3972>)
+- Update model name in rotated detection recipes
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4028>)
 
 ## \[2.2.0\]
 
 ### New features
 
 - Add RT-DETR model for Object Detection
-  (https://github.com/openvinotoolkit/training_extensions/pull/3741)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3741>)
 - Add Multi-Label & H-label Classification with torchvision models
-  (https://github.com/openvinotoolkit/training_extensions/pull/3697)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3697>)
 - Add Hugging-Face Model Wrapper for Classification
-  (https://github.com/openvinotoolkit/training_extensions/pull/3710)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3710>)
 - Add LoRA finetuning capability for ViT Architectures
-  (https://github.com/openvinotoolkit/training_extensions/pull/3729)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3729>)
 - Add Hugging-Face Model Wrapper for Object Detection
-  (https://github.com/openvinotoolkit/training_extensions/pull/3747)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3747>)
 - Add Hugging-Face Model Wrapper for Semantic Segmentation
-  (https://github.com/openvinotoolkit/training_extensions/pull/3749)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3749>)
 - Enable torch.compile to work with classification
-  (https://github.com/openvinotoolkit/training_extensions/pull/3758)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3758>)
 - Add `otx benchmark` subcommand
-  (https://github.com/openvinotoolkit/training_extensions/pull/3762)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3762>)
 - Add RTMPose for Keypoint Detection Task
-  (https://github.com/openvinotoolkit/training_extensions/pull/3781)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3781>, <https://github.com/openvinotoolkit/training_extensions/pull/4034>)
 - Add Semi-SL MeanTeacher algorithm for Semantic Segmentation
-  (https://github.com/openvinotoolkit/training_extensions/pull/3801)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3801>)
 - Update head and h-label format for hierarchical label classification
-  (https://github.com/openvinotoolkit/training_extensions/pull/3810)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3810>)
 - Support configurable input size
-  (https://github.com/openvinotoolkit/training_extensions/pull/3788)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3788>)
 - Add diffusion task
-  (https://github.com/openvinotoolkit/training_extensions/pull/3875)
+  (<https://github.com/openvinotoolkit/training_extensions/pull/3875>)
 
 ### Enhancements
 
@@ -104,9 +108,21 @@ All notable changes to this project will be documented in this file.
   (<https://github.com/openvinotoolkit/training_extensions/pull/3968>)
 - Change sematic segmentation to consider bbox only annotations
   (<https://github.com/openvinotoolkit/training_extensions/pull/3996>)
+- Relieve memory usage criteria on batch size 2 during adaptive batch size
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4009>)
+- Remove background label from RT Info for segmentation task
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4011>)
+- Prevent using too low confidence thresholds in detection
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4018>)
+- Update HPO interface
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4035>)
+- Bump onnx to 1.17.0 to omit CVE-2024-5187
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4063>)
 
 ### Bug fixes
 
+- Update anomaly base transforms to use square resizing
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4059>)
 - Fix Combined Dataloader & unlabeled warmup loss in Semi-SL
   (<https://github.com/openvinotoolkit/training_extensions/pull/3723>)
 - Revert #3579 to fix issues with replacing coco_instance with a different format in some dataset
@@ -119,6 +135,30 @@ All notable changes to this project will be documented in this file.
   (<https://github.com/openvinotoolkit/training_extensions/pull/3946>)
 - Fix config converter for tiling
   (<https://github.com/openvinotoolkit/training_extensions/pull/3973>)
+- Fix `BboxOverlaps2D` handling of empty ground-truth annotations in datasets.
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4010>)
+- Fix num_trials calculation on dataset length less than num_class
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4014>)
+- Fix out_features in HierarchicalCBAMClsHead
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4016>)
+- Fix multilabel_accuracy of MixedHLabelAccuracy
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4042>)
+- Fix wrong indices setting in HLabelInfo
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4044>)
+- Add legacy template LiteHRNet_18 template
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4049>)
+- Model templates: rename model_status value 'DISCONTINUED' to 'OBSOLETE'
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4051>)
+- Enable export of feature vectors for semantic segmentation task
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4055>)
+- Update MRCNN model export to include feature vector and saliency map
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4056>)
+- Upgrade MAPI in 2.2
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4052>)
+- Fix applying model's hparams when loading model from checkpoint
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4057>)
+- Fix incorrect all_groups order configuration in HLabelInfo
+  (<https://github.com/openvinotoolkit/training_extensions/pull/4067>)
 
 ## \[v2.1.0\]
 

README.md

@@ -197,6 +197,9 @@ In addition to the examples above, please refer to the documentation for tutoria
 - Include full image with anno in case there's no tile in tile dataset
 - Add type checker in converter for callable functions (optimizer, scheduler)
 - Change sematic segmentation to consider bbox only annotations
+- Relieve memory usage criteria on batch size 2 during adaptive batch size
+- Remove background label from RT Info for segmentation task
+- Prevent using too low confidence thresholds in detection
 
 ### Bug fixes
 
@@ -206,6 +209,10 @@ In addition to the examples above, please refer to the documentation for tutoria
 - Add missing tile recipes and various tile recipe changes
 - Change categories mapping logic
 - Fix config converter for tiling
+- Fix num_trials calculation on dataset length less than num_class
+- Fix out_features in HierarchicalCBAMClsHead
+- Fix multilabel_accuracy of MixedHLabelAccuracy
+- Fix wrong indices setting in HLabelInfo
 
 ### Known issues
 

docs/source/guide/release_notes/index.rst

@@ -4,7 +4,7 @@ Releases
 .. toctree::
   :maxdepth: 1
 
-v2.2.0 (2024.09)
+v2.2.0 (2024.10)
 ----------------
 
 New features
@@ -38,6 +38,9 @@ Enhancements
 - Include full image with anno in case there's no tile in tile dataset
 - Add type checker in converter for callable functions (optimizer, scheduler)
 - Change sematic segmentation to consider bbox only annotations
+- Relieve memory usage criteria on batch size 2 during adaptive batch size
+- Remove background label from RT Info for segmentation task
+- Prevent using too low confidence thresholds in detection
 
 Bug fixes
 ^^^^^^^^^
@@ -48,6 +51,10 @@ Bug fixes
 - Add missing tile recipes and various tile recipe changes
 - Change categories mapping logic
 - Fix config converter for tiling
+- Fix num_trials calculation on dataset length less than num_class
+- Fix out_features in HierarchicalCBAMClsHead
+- Fix multilabel_accuracy of MixedHLabelAccuracy
+- Fix wrong indices setting in HLabelInfo
 
 v2.1.0 (2024.07)
 ----------------

pyproject.toml

@@ -81,8 +81,8 @@ xpu = [
     "timm==1.0.3",
     "openvino==2024.4",
     "openvino-dev==2024.4",
-    "openvino-model-api==0.2.4",
-    "onnx==1.16.2",
+    "openvino-model-api==0.2.5",
+    "onnx==1.17.0",
     "onnxconverter-common==1.14.0",
     "nncf==2.13.0",
     "anomalib[core]==1.1.0",
@@ -96,7 +96,7 @@ base = [
     "openvino==2024.4",
     "openvino-dev==2024.4",
     "openvino-model-api==0.2.4",
-    "onnx==1.16.2",
+    "onnx==1.17.0",
     "onnxconverter-common==1.14.0",
     "nncf==2.13.0",
     "anomalib[core]==1.1.0",

src/otx/algo/classification/heads/hlabel_cls_head.py

@@ -355,7 +355,7 @@ def __init__(
         self.fc_superclass = nn.Linear(in_channels * self.step_size[0] * self.step_size[1], num_multiclass_heads)
         self.attention_fc = nn.Linear(num_multiclass_heads, in_channels * self.step_size[0] * self.step_size[1])
         self.cbam = CBAM(in_channels)
-        self.fc_subclass = nn.Linear(in_channels * self.step_size[0] * self.step_size[1], num_single_label_classes)
+        self.fc_subclass = nn.Linear(in_channels * self.step_size[0] * self.step_size[1], num_classes)
 
         self._init_layers()
 

src/otx/algo/common/losses/cross_focal_loss.py

@@ -8,7 +8,7 @@
 import torch
 import torch.nn.functional
 from torch import Tensor, nn
-from torch.cuda.amp import custom_fwd
+from torch.amp import custom_fwd
 
 from .focal_loss import py_sigmoid_focal_loss
 
@@ -79,7 +79,7 @@ def __init__(
 
         self.cls_criterion = cross_sigmoid_focal_loss
 
-    @custom_fwd(cast_inputs=torch.float32)
+    @custom_fwd(cast_inputs=torch.float32, device_type="cuda")
     def forward(
         self,
         pred: Tensor,

src/otx/algo/instance_segmentation/rtmdet_inst.py

@@ -130,6 +130,7 @@ def _exporter(self) -> OTXModelExporter:
                 "opset_version": 11,
                 "autograd_inlining": False,
             },
+            # TODO(Eugene): Add XAI support for RTMDetInst
             output_names=["bboxes", "labels", "masks", "feature_vector", "saliency_map"] if self.explain_mode else None,
         )