Neuralet.

Edge TPU
Neuralet > Model Garden > Edge TPU

Edge TPU

37 Models
Know a model that is not in this list? Submit a Model

Edge TPU

Image Classification

  • Model
  • Task
  • Inference Time (ms)
  • Throughput (FPS)
  • Description
  • File
  • Repo or Website

  • Neuralet-OFMClassifier

    Devices:

    Tasks:

    https://github.com/neuralet/neuralet-models/blob/master/edge-tpu/OFMClassifier/OFMClassifier_edgetpu.tflite

    https://github.com/neuralet/neuralet/tree/master/applications/facemask

    Official classifier by Neuralet pre-trained on the Extended-Synthetic-Blurred dataset that classifies masked faces from unmasked ones.

    30

  • InceptionV4

    Devices:

    Tasks:

    http://download.tensorflow.org/models/inception_v4_299_quant_20181026.tgz

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 299x299 that recognizes 1000 classes.

    15

  • InceptionV3

    Devices:

    Tasks:

    http://download.tensorflow.org/models/tflite_11_05_08/inception_v3_quant.tgz

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 299x299 that recognizes 1000 classes.

    18

  • InceptionV2

    Devices:

    Tasks:

    http://download.tensorflow.org/models/inception_v2_224_quant_20181026.tgz

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 224x224 that recognizes 1000 classes.

    14

  • InceptionV1

    Devices:

    Tasks:

    http://download.tensorflow.org/models/inception_v1_224_quant_20181026.tgz

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 224x224 that recognizes 1000 classes.

    14

  • MobileNetV2_iNatBirds

    Devices:

    Tasks:

    https://github.com/google-coral/edgetpu/raw/master/test_data/mobilenet_v2_1.0_224_inat_bird_quant_edgetpu.tflite

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the iNaturalist dataset (iNat birds) on an input size of 224x224 that recognizes more than 900 types of birds.

    14

  • MobileNetV2_iNatPlants

    Devices:

    Tasks:

    https://github.com/google-coral/edgetpu/raw/master/test_data/mobilenet_v2_1.0_224_inat_plant_quant_edgetpu.tflite

    https://coral.ai/models/

    Official classifier by Coral pre-trained on the iNaturalist dataset (iNat plants) on an input size of 224x224 that recognizes more than 2000 types of plants.

    13

  • EfficientNet_EdgeTpu_L

    Devices:

    Tasks:

    https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/efficientnet-edgetpu-L.tar.gz

    https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/edgetpu

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 300x300 that recognizes 1000 classes.

    12

  • EfficientNet_EdgeTpu_M

    Devices:

    Tasks:

    https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/efficientnet-edgetpu-M.tar.gz

    https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/edgetpu

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 240x240 that recognizes 1000 classes.

    13

  • EfficientNet_EdgeTpu_S

    Devices:

    Tasks:

    https://storage.googleapis.com/cloud-tpu-checkpoints/efficientnet/efficientnet-edgetpu-S.tar.gz

    https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/edgetpu

    Official classifier by Coral pre-trained on the ImageNet dataset on an input size of 224x224 that recognizes 1000 classes.

    16

Object Detection

  • Model
  • Task
  • Inference Time (ms)
  • Throughput (FPS)
  • Description
  • File
  • Repo or Website

  • SSD_MobileNetV2_OpenImagesV4

    Devices:

    Tasks:

    https://github.com/google-coral/test_data/raw/master/ssd_mobilenet_v2_face_quant_postprocess_edgetpu.tflite

    https://coral.ai/models/

    Official object detector model by Coral pre-trained on the Open Images V4 dataset on an input size of 320x320 that recognizes human face.

    19

  • SSD_MobileNetV2_COCO

    Devices:

    Tasks:

    https://github.com/google-coral/test_data/raw/master/ssd_mobilenet_v2_coco_quant_postprocess_edgetpu.tflite

    https://coral.ai/models/

    Official object detector model by Coral pre-trained on the MSCOCO dataset on an input size of 300x300 that recognizes 90 classes.

    17

  • SSD_MobileNetV1_COCO

    Devices:

    Tasks:

    https://github.com/google-coral/test_data/raw/master/ssd_mobilenet_v1_coco_quant_postprocess_edgetpu.tflite

    https://coral.ai/models/

    Official object detector model by Coral pre-trained on the MSCOCO dataset on an input size of 300x300 that recognizes 90 classes.

    17

Pose Estimation

  • Model
  • Task
  • Inference Time (ms)
  • Throughput (FPS)
  • Description
  • File
  • Repo or Website

  • PoseNet_MobileNetV1_1281x721

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/mobilenet/posenet_mobilenet_v1_075_721_1281_quant_decoder_edgetpu.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 1281x721 and a multiplier of 0.75.

    10

  • PoseNet_MobileNetV1_641x481

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/mobilenet/posenet_mobilenet_v1_075_481_641_quant_decoder_edgetpu.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 641x481 and a multiplier of 0.75.

    13

  • PoseNet_MobileNetV1_481x353

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/mobilenet/posenet_mobilenet_v1_075_353_481_quant_decoder_edgetpu.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 481x353 and a multiplier of 0.75.

    14

  • PoseNet_ResNet50_960x736

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_960_736_32_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 960x736 and an output stride of 32.

    14

  • PoseNet_ResNet50_928x672

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_928_672_16_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 928x672 and an output stride of 16.

    16

  • PoseNet_ResNet50_864x624

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_864_624_32_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 864x624 and an output stride of 32.

    12

  • PoseNet_ResNet50_768x496

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_768_496_32_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Coral pre-trained on the MSCOCO dataset on an input resolution of 768x496 and an output stride of 32.

    12

  • PoseNet_ResNet50_640x480

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_640_480_16_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Google pre-trained on the MSCOCO dataset on an input resolution of 640x480 and an output stride of 16.

    12

  • PoseNet_ResNet50_416x288

    Devices:

    Tasks:

    https://github.com/google-coral/project-posenet/blob/master/models/resnet/posenet_resnet_50_416_288_16_quant_edgetpu_decoder.tflite

    https://github.com/neuralet/neuralet-models

    Official pose estimation model by Google pre-trained on the MSCOCO dataset on an input resolution of 416x288 and an output stride of 16.

    11

Semantic Segmentation

  • Model
  • Task
  • Inference Time (ms)
  • Throughput (FPS)
  • Description
  • File
  • Repo or Website

  • UNet_MobileNetV2_256x256

    Devices:

    Tasks:

    https://github.com/google-coral/test_data/raw/master/keras_post_training_unet_mv2_256_quant_edgetpu.tflite

    https://coral.ai/models/

    Official model by Coral that recognizes and segments pets using 3 classes: pixels belonging to a pet, pixels bordering a pet, and background pixels (it does not classify the type of pet), and is pre-trained on the Oxford-IIIT Pet dataset on an input size of 256x256.

    13

  • UNet_MobileNetV2_128x128

    Devices:

    Tasks:

    https://github.com/google-coral/test_data/raw/master/keras_post_training_unet_mv2_128_quant_edgetpu.tflite

    https://coral.ai/models/

    Official model by Coral that recognizes and segments pets using 3 classes: pixels belonging to a pet, pixels bordering a pet, and background pixels (it does not classify the type of pet), and is pre-trained on the Oxford-IIIT Pet dataset on an input size of 128x128.

    13

  • DeepLabV3_MobileNetV2_513x513_1.0

    Devices:

    Tasks:

    https://github.com/google-coral/edgetpu/raw/master/test_data/deeplabv3_mnv2_pascal_quant_edgetpu.tflite

    https://coral.ai/models/

    Official model by Coral that recognizes and segments 20 types of objects and is pre-trained on the PASCAL VOC 2012 dataset on an input size of 513x513 and depth multiplier of 1.0.

    12

  • DeepLabV3_MobileNetV2_513x513_0.5

    Devices:

    Tasks:

    https://github.com/google-coral/edgetpu/raw/master/test_data/deeplabv3_mnv2_dm05_pascal_quant_edgetpu.tflite

    https://coral.ai/models/

    Official model by Coral that recognizes and segments 20 types of objects and is pre-trained on the PASCAL VOC 2012 dataset on an input size of 513x513 and depth multiplier of 0.5.

    16

  • BodyPix_ResNet50_960x736

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_960_736_32_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 960x736 and output stride of 32.

    11

  • BodyPix_ResNet50_928x672

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_928_672_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 928x672 and output stride of 16.

    10

  • BodyPix_ResNet50_864x624

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_864_624_32_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 864x624 and output stride of 32.

    11

  • BodyPix_ResNet50_768x496

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_768_496_32_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 768x496 and output stride of 32.

    10

  • BodyPix_ResNet50_640x480

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_640_480_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 640x480, depth multiplier of 0.75, and output stride of 16.

    11

  • BodyPix_ResNet50_416x288

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_resnet_50_416_288_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 416x288, depth multiplier of 0.75, and output stride of 16.

    11

  • BodyPix_MobileNetV1_768x576

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_mobilenet_v1_075_768_576_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 768x576, depth multiplier of 0.75, and output stride of 16.

    10

  • BodyPix_MobileNetV1_640x480

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_mobilenet_v1_075_640_480_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 640x480, depth multiplier of 0.75, and output stride of 16.

    11

  • BodyPix_MobileNetV1_480x352

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_mobilenet_v1_075_480_352_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 480x352, depth multiplier of 0.75, and output stride of 16.

    12

  • BodyPix_MobileNetV1_1280x720

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_mobilenet_v1_075_1280_720_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 1280x720, depth multiplier of 0.75, and output stride of 16.

    9

  • BodyPix_MobileNetV1_1024x768

    Devices:

    Tasks:

    https://github.com/google-coral/project-bodypix/blob/master/models/bodypix_mobilenet_v1_075_1024_768_16_quant_edgetpu_decoder.tflite

    https://coral.ai/models/

    Official model by Coral that allows for person and body-part segmentation and is pre-trained with an input size of 1024x768, depth multiplier of 0.75, and output stride of 16.

    16

Get In Touch

Have a question? Send us a message and we will respond as soon as possible.