Vitis AI Library 1.4 Release Notes - 1.4.1 English

Vitis AI Library User Guide (UG1354)

Document ID
UG1354
Release Date
2021-12-11
Version
1.4.1 English

This section contains information regarding the features and updates of the Vitis™ AI Library 1.4 release.

Key Features And Enhancements

This Vitis AI Library release includes the following key features and enhancements:

New Board Support
SoM Series: KV260 is supported in this release.
New DPU Support

Enable new DPU target DPUCVDX8G designed for VCK190.

New Model Libraries
The following new model libraries are supported.
New Model Support
  • Added ten new PyTorch models
  • Added five new TensorFlow models, including one TensorFlow2 model.
  • Added one new Caffe model
New Deploy APIs Support

graph_runner is introduced for deploying model.

New Tool Support

xdputil is introduced for the dpu and xmodel debug.

Changes

  • None.

Compatibility

The Vitis™ AI Library 1.4 is tested with the following images.

  • xilinx-zcu102-dpu-v2021.1-v1.4.0.img.gz
  • xilinx-zcu104-dpu-v2021.1-v1.4.0.img.gz
  • xilinx-kv260-dpu-v2020.2-v1.4.0.img.gz
  • xilinx-vck190-dpu-v2020.2-v1.4.0.img.gz

Model Support

The following models are supported by this version of the Vitis™ AI Library.

Table 1. Models Supported by the AI Library
No. Neural Network ZCU102/

ZCU104/

KV260
VCK190 U50/U50LV/U280-DPUCAHX8H U50/U50LV/U280-DPUCAHX8L VCK5000-DPUCVDX8H Application
1 inception_resnet_v2_tf Y Y Y Y N/A Image Classification
2 inception_v1_tf Y Y Y Y Y
3 inception_v3_tf Y Y Y Y N/A
4 inception_v4_2016_09_09_tf Y Y Y Y N/A
5 mobilenet_v1_0_25_128_tf Y N/A N/A N/A N/A
6 mobilenet_v1_0_5_160_tf Y N/A N/A Y N/A
7 mobilenet_v1_1_0_224_tf Y N/A N/A Y N/A
8 mobilenet_v2_1_0_224_tf Y N/A N/A Y N/A
9 mobilenet_v2_1_4_224_tf Y N/A N/A Y N/A
10 resnet_v1_101_tf Y Y Y Y Y
11 resnet_v1_152_tf Y Y Y Y Y
12 resnet_v1_50_tf Y Y Y Y Y
13 vgg_16_tf Y Y Y Y N/A
14 vgg_19_tf Y Y Y Y N/A
15 ssd_mobilenet_v1_coco_tf Y N/A N/A Y N/A Object Detection
16 ssd_mobilenet_v2_coco_tf Y N/A N/A Y N/A
17 ssd_resnet_50_fpn_coco_tf Y Y Y N/A N/A
18 yolov3_voc_tf Y Y Y N/A Y
19 mlperf_ssd_resnet34_tf Y Y Y Y N/A
20 resnet50 Y Y Y Y Y Image Classification
21 resnet18 Y Y Y Y Y
22 inception_v1 Y Y Y Y Y
23 inception_v2 Y Y Y Y N/A
24 inception_v3 Y Y Y Y N/A
25 inception_v4 Y Y Y Y N/A
26 mobilenet_v2 Y N/A N/A Y N/A
27 squeezenet Y Y Y Y Y
28 ssd_pedestrian_pruned_0_97 Y Y Y Y N/A ADAS Pedestrian Detection
29 ssd_traffic_pruned_0_9 Y Y Y Y Y Traffic Detection
30 ssd_adas_pruned_0_95 Y Y Y Y Y ADAS Vehicle Detection
31 ssd_mobilenet_v2 Y N/A N/A Y N/A Object Detection
32 refinedet_pruned_0_8 Y Y Y Y Y
33 refinedet_pruned_0_92 Y Y Y Y Y
34 refinedet_pruned_0_96 Y Y Y Y Y
35 vpgnet_pruned_0_99 Y Y Y Y Y ADAS Lane Detection
36 fpn Y Y Y Y Y ADAS Segmentation
37 sp_net Y Y Y Y Y Pose Estimation
38 openpose_pruned_0_3 Y Y Y Y Y
39 densebox_320_320 Y Y Y N/A Y Face Detection
40 densebox_640_360 Y Y Y N/A Y
41 face_landmark Y Y Y Y Y Face Detection and Recognition
42 reid Y Y Y Y Y Object tracking
43 multi_task Y Y Y Y Y ADAS
44 yolov3_adas_pruned_0_9 Y Y Y N/A Y Object Detection
45 yolov3_voc Y Y Y N/A Y
46 yolov3_bdd Y Y Y N/A Y
47 yolov2_voc Y Y Y N/A Y
48 yolov2_voc_pruned_0_66 Y Y Y N/A Y
49 yolov2_voc_pruned_0_71 Y Y Y N/A Y
50 yolov2_voc_pruned_0_77 Y Y Y N/A Y
51 facerec_resnet20 Y Y Y Y N/A Face Recognition
52 facerec_resnet64 Y Y Y Y N/A
53 plate_detection Y Y Y Y Y Plate Recognition
54 plate_recognition Y Y Y N/A N/A
55 FPN_Res18_Medical_segmentation Y Y Y Y Y Medical Segmentation
56 refinedet_baseline Y Y Y Y Y Object Detection
57 resnet50_pt Y Y Y Y Y Image Classification
58 squeezenet_pt Y Y Y Y Y
59 inception_v3_pt Y Y Y Y N/A
60

personreid-res50_pt

Y Y Y Y N/A Object Tracking
61

facereid-large_pt

Y Y Y N/A N/A
62

facereid-small_pt

Y Y Y Y N/A
63

SemanticFPN_cityscapes_pt

Y Y Y Y Y Segmentation
64

facerec-resnet20_mixed_pt

Y Y Y Y N/A Face Recognition
65 face-quality_pt Y Y Y Y Y
66 MT-resnet18_mixed_pt Y N/A N/A N/A N/A ADAS
67 salsanext_pt Y Y Y Y Y Point Cloud
68 pointpillars_kitti_12000_0_pt

pointpillars_kitti_12000_1_pt

Y N/A N/A N/A N/A
69 unet_chaos-CT_pt Y Y Y N/A Y CT Segmentation
70 FPN-resnet18_covid19-seg_pt Y Y Y Y Y Covid-19 Segmentation
71 ENet_cityscapes_pt Y Y Y Y Y Segmentation
72 personreid-res18_pt Y Y Y N/A N/A Object Tracking
73 yolov4_leaky_spp_m Y Y Y N/A N/A Object Detection
74 hourglass-pe_mpii Y N/A N/A N/A N/A Pose Estimation
75 retinaface Y N/A N/A N/A N/A Face Detection
76 FPN-resnet18_Endov Y N/A N/A N/A N/A Robot Instrument Segmentation
77 tiny_yolov3_vmss Y Y Y N/A Y Object Detection
78 face-quality Y Y Y Y N/A Face Recognition
79 ssdlite_mobilenet_v2_coco_tf Y N/A N/A Y N/A Object Detection
80 ssd_inception_v2_coco_tf Y N/A N/A N/A N/A
81 MLPerf_resnet50_v1.5_tf Y Y Y Y Y Image Classification
82 mobilenet_edge_1_0_tf Y N/A N/A N/A N/A
83 mobilenet_edge_0_75_tf Y N/A N/A N/A N/A
84 refinedet_VOC_tf Y Y Y Y Y Object Detection
85 RefineDet-Medical_EDD_tf Y Y Y Y Y Medical Detection
86 resnet_v2_50_tf Y N/A N/A N/A N/A Image Classification
87 resnet_v2_101_tf Y N/A N/A N/A N/A
88 resnet_v2_152_tf Y N/A N/A N/A N/A
89 mobilenet_v2_cityscapes_tf Y N/A N/A N/A N/A Segmentation
90 inception_v2_tf Y N/A N/A Y N/A Image Classification
91 resnet50_tf2 Y Y Y Y Y
92 mobilenet_1_0_224_tf2 Y N/A N/A Y N/A
93 inception_v3_tf2 Y Y Y Y N/A
94 medical_seg_cell_tf2 Y Y Y Y Y Medical Segmentation
95 semantic_seg_citys_tf2 Y Y Y Y N/A Segmentation
96 efficientNet-edgetpu-S_tf Y Y N/A N/A N/A Image Classification
97 efficientNet-edgetpu-M_tf Y Y N/A N/A N/A
98 efficientNet-edgetpu-L_tf Y Y N/A N/A N/A
99 SemanticFPN_Mobilenetv2_pt Y Y N/A Y N/A Segmentation
100 pointpillars_nuscenes_40000_64_0_pt

pointpillars_nuscenes_40000_64_1_pt

Y Y N/A N/A N/A 3D object detection
101 pointpainting_nuscenes_40000_64_0_pt

pointpainting_nuscenes_40000_64_1_pt

Y Y N/A N/A N/A 2D semantic segmentation and 3D object detection
102 salsanext_v2_pt Y Y Y N/A N/A 3D Segmentation
103 centerpoint_0_pt

centerpoint_1_pt

Y Y N/A N/A N/A 4D radar detection
104 multi_task_v3_pt Y Y N/A N/A N/A ADAS
105 FADNet_0_pt

FADNet_1_pt

FADNet_2_pt

Y Y N/A N/A N/A Depth Estimation
106 rcan_pruned_tf Y Y N/A Y N/A Super Resolution
107 efficientnet_tf N/A Y N/A N/A N/A Classification
108 yolov4_leaky_spp_m_pruned_0_36 Y Y Y N/A N/A Object Detection
109 pmg_pt Y Y Y N/A N/A Brand Recognition
110 bbc_pt Y Y N/A N/A N/A Bayesian Crowd Counting
111 SA_gate_pt N/A Y N/A N/A N/A Indoor Segmentation
  1. Networks with the suffix "_tf" or "_tf2" were trained on TensorFlow.
  2. Networks with the suffix "_pt" were trained on PyTorch.
  3. Networks with no suffix were trained on Caffe.
  4. The column of U50-V3ME/U50LV-V3ME/U280-V3ME is for DPUCAHX8L.

Device Support

The following platforms and evaluation boards (EVB) are supported by the Vitis™ AI Library 1.4.

Table 2. Edge Device Support
Platform EVB Version
Zynq UltraScale+ MPSoC ZU9EG Xilinx ZCU102 1.1
Zynq® UltraScale+™ MPSoC ZU7EV Xilinx ZCU104 1.0
Zynq UltraScale+ MPSoC Xilinx Kria KV260 1.0
Versal AI Core series VC1902 Xilinx VCK190 ES1
Table 3. Cloud Board Support
Accelerator Cards
Xilinx Alveo U50 Data Center accelerator card
Xilinx Alveo U50LV Data Center accelerator card
Xilinx Alveo U200 Data Center accelerator card
Xilinx Alveo U250 Data Center accelerator card
Xilinx Alveo U280 Data Center accelerator card
Versal AI Core series VCK5000 Data Center development kit

Limitations

  • Some neural networks with MobileNet as the backbone are not supported on the Alveo U50, U50LV, and U280 accelerator cards.
  • Due to limitations of the Docker environment, Multi-task demos cannot run on the DRM mode on Cloud boards.