Skip to content

Latest commit

 

History

History
230 lines (185 loc) · 28.6 KB

validation_results_tflite.md

File metadata and controls

230 lines (185 loc) · 28.6 KB

Validation results for the models inferring using TensorFlow Lite

Public models (Open Model Zoo)

Image classification

Test image #1

Data source: ImageNet

Image resolution: 709 x 510 

Model Parameters Python API
densenet-121 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.9523314 Granny Smith
0.0132282 orange
0.0125180 lemon
0.0027912 banana
0.0020333 piggy bank, penny bank
googlenet-v1-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.6735930 Granny Smith
0.0737855 piggy bank, penny bank
0.0155380 vase
0.0154004 pitcher, ewer
0.0136552 saltshaker, salt shaker
googlenet-v2-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.9849940 Granny Smith
0.0010004 lemon
0.0009706 pomegranate
0.0006835 tennis ball
0.0006694 banana
googlenet-v3 Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.9867674 Granny Smith
0.0008529 bikini, two-piece
0.0005354 piggy bank, penny bank
0.0003701 pomegranate
0.0001682 pool table, billiard table, snooker table
googlenet-v4-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.9969806 Granny Smith
0.0001207 Rhodesian ridgeback
0.0000488 hair slide
0.0000473 pineapple, ananas
0.0000330 banana
mobilenet-v1-1.0-224-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.2949494 pitcher, ewer
0.1867124 saltshaker, salt shaker
0.1249271 necklace
0.0867643 piggy bank, penny bank
0.0360211 Granny Smith
mobilenet-v2-1.0-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.4164807 Granny Smith
0.3500757 piggy bank, penny bank
0.0358796 saltshaker, salt shaker
0.0147685 vase
0.0131548 pitcher, ewer
mobilenet-v2-1.4-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.3428614 saltshaker, salt shaker
0.0935006 vase
0.0899924 Granny Smith
0.0667358 pitcher, ewer
0.0666182 piggy bank, penny bank
inception-resnet-v2-tf Image resolution: 299x299.
Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
8.0107708 Granny Smith
4.4552484 piggy bank, penny bank
4.2636762 bell pepper
3.9343853 candle, taper, wax light
3.5531902 pomegranate
mobilenet-v3-small-1.0-224 - 0.4481893 Granny Smith
0.0884615 lemon
0.0727510 pop bottle, soda bottle
0.0331238 saltshaker, salt shaker
0.0218442 pitcher, ewer
mobilenet-v3-large-1.0-224 - 0.6718515 Granny Smith
0.1939126 piggy bank, penny bank
0.0254287 lemon
0.0245753 vase
0.0090322 teapot
resnet-50-tf Mean: [123.675,116.28,103.53]. 0.9553038 Granny Smith
0.0052123 lemon
0.0047185 piggy bank, penny bank
0.0045875 orange
0.0044233 necklace

Test image #2

Data source: ImageNet

Image resolution: 500 x 500 

Model Parameters Python API
densenet-121 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.9841611 junco, snowbird
0.0072198 chickadee
0.0034962 brambling, Fringilla montifringilla
0.0016226 water ouzel, dipper
0.0012858 indigo bunting, indigo finch, indigo bird, Passerina cyanea
googlenet-v1-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.7443183 junco, snowbird
0.0474523 brambling, Fringilla montifringilla
0.0457429 chickadee
0.0213391 goldfinch, Carduelis carduelis
0.0085102 house finch, linnet, Carpodacus mexicanus
googlenet-v2-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.9265908 junco, snowbird
0.0166747 brambling, Fringilla montifringilla
0.0058714 chickadee
0.0026126 water ouzel, dipper
0.0022344 indigo bunting, indigo finch, indigo bird, Passerina cyanea
googlenet-v3 Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.9488292 junco, snowbird
0.0005887 water ouzel, dipper
0.0004797 iron, smoothing iron
0.0003071 indigo bunting, indigo finch, indigo bird, Passerina cyanea
0.0002692 cleaver, meat cleaver, chopper
googlenet-v4-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.9339716 junco, snowbird
0.0006892 chickadee
0.0005481 brambling, Fringilla montifringilla
0.0004948 indigo bunting, indigo finch, indigo bird, Passerina cyanea
0.0004539 water ouzel, dipper
mobilenet-v1-1.0-224-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.7099941 junco, snowbird
0.2239839 chickadee
0.0195020 goldfinch, Carduelis carduelis
0.0140457 jay
0.0136091 brambling, Fringilla montifringilla
mobilenet-v2-1.0-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.3981952 junco, snowbird
0.0649636 chickadee
0.0456628 brambling, Fringilla montifringilla
0.0063850 water ouzel, dipper
0.0041957 goldfinch, Carduelis carduelis
mobilenet-v2-1.4-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.7363465 chickadee
0.0283495 junco, snowbird
0.0117877 brambling, Fringilla montifringilla
0.0083691 goldfinch, Carduelis carduelis
0.0035830 water ouzel, dipper
inception-resnet-v2-tf Image resolution: 299x299.
Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
10.0273972 junco, snowbird
4.6770372 brambling, Fringilla montifringilla
4.2079940 goldfinch, Carduelis carduelis
4.1425276 water ouzel, dipper
4.0244055 chickadee
mobilenet-v3-small-1.0-224 - 0.5813942 junco, snowbird
0.0588930 brambling, Fringilla montifringilla
0.0446762 house finch, linnet, Carpodacus mexicanus
0.0411857 goldfinch, Carduelis carduelis
0.0150912 chickadee
mobilenet-v3-large-1.0-224 - 0.7943738 junco, snowbird
0.0318200 brambling, Fringilla montifringilla
0.0084637 water ouzel, dipper
0.0071047 goldfinch, Carduelis carduelis
0.0061734 chickadee
resnet-50-tf Mean: [123.675,116.28,103.53]. 0.9983400 junco, snowbird
0.0004680 brambling, Fringilla montifringilla
0.0003848 chickadee
0.0003656 water ouzel, dipper
0.0003383 goldfinch, Carduelis carduelis

Test image #3

Data source: ImageNet

Image resolution: 333 x 500 

Model Parameters Python API
densenet-121 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.3022473 liner, ocean liner
0.1322417 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.1194588 container ship, containership, container vessel
0.0795097 drilling platform, offshore rig
0.0723070 dock, dockage, docking facility
googlenet-v1-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.1235979 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.1017586 liner, ocean liner
0.0949444 drilling platform, offshore rig
0.0817947 container ship, containership, container vessel
0.0486889 fireboat
googlenet-v2-tf Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.2662660 container ship, containership, container vessel
0.0966037 dock, dockage, docking facility
0.0876837 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0488674 beacon, lighthouse, beacon light, pharos
0.0343599 drilling platform, offshore rig
googlenet-v3 Mean: [127.5, 127.5, 127.5].
Input scale: [127.5].
0.4653829 beacon, lighthouse, beacon light, pharos
0.3437532 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0512180 submarine, pigboat, sub, U-boat
0.0174647 liner, ocean liner
0.0134649 lifeboat
googlenet-v4-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.6737013 lifeboat
0.0432948 submarine, pigboat, sub, U-boat
0.0322841 fireboat
0.0264144 beacon, lighthouse, beacon light, pharos
0.0147488 drilling platform, offshore rig
mobilenet-v1-1.0-224-tf Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.1175058 lifeboat
0.1106691 aircraft carrier, carrier, flattop, attack aircraft carrier
0.1055247 liner, ocean liner
0.0836357 beacon, lighthouse, beacon light, pharos
0.0784211 drilling platform, offshore rig
mobilenet-v2-1.0-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.2761748 beacon, lighthouse, beacon light, pharos
0.1192475 liner, ocean liner
0.0864237 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0541655 drilling platform, offshore rig
0.0266723 container ship, containership, container vessel
mobilenet-v2-1.4-224 Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
0.2250047 beacon, lighthouse, beacon light, pharos
0.2051269 container ship, containership, container vessel
0.1319712 liner, ocean liner
0.0256291 dock, dockage, docking facility
0.0241968 breakwater, groin, groyne, mole, bulwark, seawall, jetty
inception-resnet-v2-tf Image resolution: 299x299.
Mean: [123.675,116.28,103.53].
Input scale: [58.395,57.12,57.375].
6.4228325 breakwater, groin, groyne, mole, bulwark, seawall, jetty
6.0842223 liner, ocean liner
5.8280630 fireboat
5.7098336 dock, dockage, docking facility
5.6666737 container ship, containership, container vessel
mobilenet-v3-small-1.0-224 - 0.0980954 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0957138 container ship, containership, container vessel
0.0853775 pirate, pirate ship
0.0690932 drilling platform, offshore rig
0.0685616 lifeboat
mobilenet-v3-large-1.0-224 - 0.1806492 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.1449896 lifeboat
0.1315165 submarine, pigboat, sub, U-boat
0.0884149 dock, dockage, docking facility
0.0476540 fireboat
resnet-50-tf Mean: [123.675,116.28,103.53]. 0.2357713 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.1480762 liner, ocean liner
0.1104688 container ship, containership, container vessel
0.1095407 drilling platform, offshore rig
0.0915569 beacon, lighthouse, beacon light, pharos

Other tasks

[TBD]

Public models (TF hub)

Image classification

Test image #1

Data source: ImageNet

Image resolution: 709 x 510 

Model Parameters Python API
mobilenet_v1_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.7150755 Granny Smith
0.0202576 piggy bank, penny bank
0.0088377 teapot
0.0072254 bell pepper
0.0058900 banana
mobilenet_v2_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.6549551 Granny Smith
0.1130055 piggy bank, penny bank
0.0566443 teapot
0.0250644 saltshaker, salt shaker
0.0120769 vase
mobilenet_v3_small_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.5426094 Granny Smith
0.0725947 teapot
0.0285967 piggy bank, penny bank
0.0269885 pitcher, ewer
0.0195926 vase
mobilenet_v3_large_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.8494291 Granny Smith
0.0085453 piggy bank, penny bank
0.0069065 lemon
0.0055464 tennis ball
0.0027605 pomegranate
mobilenet_v1_100_224_uint8_1 - 248.0000000 Granny Smith
1.0000000 lemon
0.0000000 sulphur butterfly, sulfur butterfly
0.0000000 guinea pig, Cavia cobaya
0.0000000 beaver
mobilenet_v2_100_224_uint8_1 - 242.0000000 Granny Smith
2.0000000 piggy bank, penny bank
1.0000000 lemon
1.0000000 teapot
1.0000000 saltshaker, salt shaker
mobilenet_v3_small_100_224_uint8_1 - 25.0000000 junco, snowbird
13.0000000 water ouzel, dipper
11.0000000 chickadee
8.0000000 brambling, Fringilla montifringilla
7.0000000 goldfinch, Carduelis carduelis
mobilenet_v3_large_100_224_uint8_1 - 204.0000000 Granny Smith
6.0000000 piggy bank, penny bank
1.0000000 orange
1.0000000 teapot
1.0000000 vase
efficientnet_lite0_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1806599 dumbbell
0.1265180 screw
0.0506000 Granny Smith
0.0467094 barbell
0.0345199 vase
efficientnet_lite1_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.0281865 wool, woolen, woollen
0.0269535 water bottle
0.0259669 trilobite
0.0250501 Granny Smith
0.0229548 teapot
efficientnet_lite2_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1039253 vase
0.1024296 pitcher, ewer
0.0350566 teapot
0.0325848 hook, claw
0.0292138 cup
efficientnet_lite3_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1926491 pitcher, ewer
0.1502650 teapot
0.1470646 electric fan, blower
0.1428124 water jug
0.1257372 strainer0.1926491 pitcher, ewer
0.1502650 teapot
0.1470646 electric fan, blower
0.1428124 water jug
0.1257372 strainer
efficientnet_lite4_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.5777039 Granny Smith
0.1232639 teapot
0.0563486 pitcher, ewer
0.0132013 jack-o'-lantern
0.0130741 orange
lite-efficientnet_lite0_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
81.0000000 Granny Smith
30.0000000 bell pepper
27.0000000 orange
17.0000000 pomegranate
10.0000000 strawberry
lite-efficientnet_lite1_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
25.0000000 bell pepper
20.0000000 Granny Smith
18.0000000 orange
16.0000000 candle, taper, wax light
11.0000000 piggy bank, penny bank
lite-efficientnet_lite2_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
39.0000000 Granny Smith
27.0000000 orange
22.0000000 piggy bank, penny bank
13.0000000 bell pepper
12.0000000 candle, taper, wax light
lite-efficientnet_lite3_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
207.0000000 Granny Smith
10.0000000 tennis ball
9.0000000 orange
4.0000000 piggy bank, penny bank
3.0000000 lemon
lite-efficientnet_lite4_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
141.0000000 Granny Smith
19.0000000 orange
5.0000000 piggy bank, penny bank
4.0000000 banana
3.0000000 jack-o'-lantern
efficientnet_lite0_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
81.0000000 Granny Smith
30.0000000 bell pepper
27.0000000 orange
17.0000000 pomegranate
10.0000000 strawberry
efficientnet_lite1_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
25.0000000 bell pepper
20.0000000 Granny Smith
18.0000000 orange
16.0000000 candle, taper, wax light
11.0000000 piggy bank, penny bank
efficientnet_lite2_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
39.0000000 Granny Smith
27.0000000 orange
22.0000000 piggy bank, penny bank
13.0000000 bell pepper
12.0000000 candle, taper, wax light
efficientnet_lite3_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
207.0000000 Granny Smith
10.0000000 tennis ball
9.0000000 orange
4.0000000 piggy bank, penny bank
3.0000000 lemon
efficientnet_lite4_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
141.0000000 Granny Smith
19.0000000 orange
5.0000000 piggy bank, penny bank
4.0000000 banana
3.0000000 jack-o'-lantern

Test image #2

Data source: ImageNet

Image resolution: 500 x 500 

Model Parameters Python API
mobilenet_v1_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.8584890 junco, snowbird
0.0235178 brambling, Fringilla montifringilla
0.0185745 goldfinch, Carduelis carduelis
0.0094353 water ouzel, dipper
0.0067247 chickadee
mobilenet_v2_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.8474838 junco, snowbird
0.0402333 chickadee
0.0112412 brambling, Fringilla montifringilla
0.0056867 water ouzel, dipper
0.0020902 goldfinch, Carduelis carduelis
mobilenet_v3_small_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.6117089 junco, snowbird
0.0544940 chickadee
0.0274826 goldfinch, Carduelis carduelis
0.0213195 water ouzel, dipper
0.0100568 brambling, Fringilla montifringilla
mobilenet_v3_large_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.6303865 junco, snowbird
0.0645584 brambling, Fringilla montifringilla
0.0283644 goldfinch, Carduelis carduelis
0.0097315 chickadee
0.0092374 water ouzel, dipper
mobilenet_v1_100_224_uint8_1 - 181.0000000 junco, snowbird
20.0000000 brambling, Fringilla montifringilla
12.0000000 goldfinch, Carduelis carduelis
3.0000000 chickadee
3.0000000 house finch, linnet, Carpodacus mexicanus
mobilenet_v2_100_224_uint8_1 - 221.0000000 junco, snowbird
6.0000000 chickadee
3.0000000 brambling, Fringilla montifringilla
1.0000000 water ouzel, dipper
0.0000000 toilet tissue, toilet paper, bathroom tissue
mobilenet_v3_small_100_224_uint8_1 - 97.0000000 Granny Smith
27.0000000 tennis ball
8.0000000 ping-pong ball
5.0000000 candle, taper, wax light
5.0000000 saltshaker, salt shaker
mobilenet_v3_large_100_224_uint8_1 - 144.0000000 junco, snowbird
9.0000000 goldfinch, Carduelis carduelis
6.0000000 water ouzel, dipper
5.0000000 chickadee
5.0000000 brambling, Fringilla montifringilla
efficientnet_lite0_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.0890195 chickadee
0.0884429 bee eater
0.0799505 goldfinch, Carduelis carduelis
0.0760872 brambling, Fringilla montifringilla
0.0670157 hummingbird
efficientnet_lite1_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1486115 guenon, guenon monkey
0.0709516 three-toed sloth, ai, Bradypus tridactylus
0.0540262 marmoset
0.0501767 toucan
0.0370303 junco, snowbird
efficientnet_lite2_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.2191649 bulbul
0.2073708 three-toed sloth, ai, Bradypus tridactylus
0.1857240 chickadee
0.0258803 junco, snowbird
0.0256731 house finch, linnet, Carpodacus mexicanus
efficientnet_lite3_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.4806626 junco, snowbird
0.1389775 brambling, Fringilla montifringilla
0.1015250 house finch, linnet, Carpodacus mexicanus
0.0838527 chickadee
0.0206517 bulbul
efficientnet_lite4_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.3836534 chickadee
0.3339088 junco, snowbird
0.0034993 brambling, Fringilla montifringilla
0.0031002 water ouzel, dipper
0.0028740 hummingbird
lite-efficientnet_lite0_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
38.0000000 bee eater
31.0000000 chickadee
31.0000000 house finch, linnet, Carpodacus mexicanus
29.0000000 junco, snowbird
24.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
lite-efficientnet_lite1_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
84.0000000 jay
28.0000000 junco, snowbird
11.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
9.0000000 chickadee
7.0000000 little blue heron, Egretta caerulea
lite-efficientnet_lite2_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
73.0000000 junco, snowbird
26.0000000 chickadee
8.0000000 brambling, Fringilla montifringilla
4.0000000 bulbul
3.0000000 jay
lite-efficientnet_lite3_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
58.0000000 jay
26.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
20.0000000 chickadee
17.0000000 jacamar
11.0000000 bee eater
lite-efficientnet_lite4_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
150.0000000 hummingbird
20.0000000 junco, snowbird
11.0000000 jacamar
6.0000000 water ouzel, dipper
4.0000000 chickadee
efficientnet_lite0_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
38.0000000 bee eater
31.0000000 chickadee
31.0000000 house finch, linnet, Carpodacus mexicanus
29.0000000 junco, snowbird
24.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
efficientnet_lite1_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
84.0000000 jay
28.0000000 junco, snowbird
11.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
9.0000000 chickadee
7.0000000 little blue heron, Egretta caerulea
efficientnet_lite2_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
73.0000000 junco, snowbird
26.0000000 chickadee
8.0000000 brambling, Fringilla montifringilla
4.0000000 bulbul
3.0000000 jay
efficientnet_lite3_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
58.0000000 jay
26.0000000 indigo bunting, indigo finch, indigo bird, Passerina cyanea
20.0000000 chickadee
17.0000000 jacamar
11.0000000 bee eater
efficientnet_lite4_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
150.0000000 hummingbird
20.0000000 junco, snowbird
11.0000000 jacamar
6.0000000 water ouzel, dipper
4.0000000 chickadee

Test image #3

Data source: ImageNet

Image resolution: 333 x 500 

Model Parameters Python API
mobilenet_v1_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.1904013 liner, ocean liner
0.0967771 lifeboat
0.0881745 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0482143 beacon, lighthouse, beacon light, pharos
0.0478232 catamaran
mobilenet_v2_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.0606234 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0568781 container ship, containership, container vessel
0.0518561 lifeboat
0.0431800 beacon, lighthouse, beacon light, pharos
0.0337389 drilling platform, offshore rig
mobilenet_v3_small_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.0724729 aircraft carrier, carrier, flattop, attack aircraft carrier
0.0653515 drilling platform, offshore rig
0.0644393 liner, ocean liner
0.0580560 container ship, containership, container vessel
0.0499799 beacon, lighthouse, beacon light, pharos
mobilenet_v3_large_100_224_fp32_1 Mean: [127.5,127.5,127.5].
Input scale: [127.5,127.5,127.5].
0.2661534 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0744180 beacon, lighthouse, beacon light, pharos
0.0523450 container ship, containership, container vessel
0.0406797 pirate, pirate ship
0.0398768 aircraft carrier, carrier, flattop, attack aircraft carrier
mobilenet_v1_100_224_uint8_1 - 34.0000000 catamaran
26.0000000 liner, ocean liner
21.0000000 water bottle
10.0000000 dock, dockage, docking facility
9.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
mobilenet_v2_100_224_uint8_1 - 15.0000000 pirate, pirate ship
14.0000000 liner, ocean liner
7.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
7.0000000 container ship, containership, container vessel
6.0000000 catamaran
mobilenet_v3_small_100_224_uint8_1 - 25.0000000 drilling platform, offshore rig
17.0000000 beacon, lighthouse, beacon light, pharos
15.0000000 aircraft carrier, carrier, flattop, attack aircraft carrier
12.0000000 container ship, containership, container vessel
12.0000000 sandbar, sand bar
mobilenet_v3_large_100_224_uint8_1 - 67.0000000 beacon, lighthouse, beacon light, pharos
40.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
32.0000000 container ship, containership, container vessel
10.0000000 fireboat
9.0000000 seashore, coast, seacoast, sea-coast
efficientnet_lite0_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1718010 liner, ocean liner
0.1557053 aircraft carrier, carrier, flattop, attack aircraft carrier
0.1440452 submarine, pigboat, sub, U-boat
0.0589084 beacon, lighthouse, beacon light, pharos
0.0228401 space shuttle
efficientnet_lite1_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.1246776 beacon, lighthouse, beacon light, pharos
0.0895273 liner, ocean liner
0.0312974 aircraft carrier, carrier, flattop, attack aircraft carrier
0.0166320 submarine, pigboat, sub, U-boat
0.0145705 breakwater, groin, groyne, mole, bulwark, seawall, jetty
efficientnet_lite2_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.2495511 beacon, lighthouse, beacon light, pharos
0.1150195 submarine, pigboat, sub, U-boat
0.0678146 liner, ocean liner
0.0312395 breakwater, groin, groyne, mole, bulwark, seawall, jetty
0.0109627 catamaran
efficientnet_lite3_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.0830593 liner, ocean liner
0.0747621 beacon, lighthouse, beacon light, pharos
0.0370711 container ship, containership, container vessel
0.0300901 sewing machine
0.0277146 breakwater, groin, groyne, mole, bulwark, seawall, jetty
efficientnet_lite4_fp32_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
0.9409395 liner, ocean liner
0.0027906 container ship, containership, container vessel
0.0024161 dock, dockage, docking facility
0.0019609 fireboat
0.0011747 breakwater, groin, groyne, mole, bulwark, seawall, jetty
lite-efficientnet_lite0_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
53.0000000 lifeboat
34.0000000 catamaran
23.0000000 speedboat
15.0000000 liner, ocean liner
12.0000000 fireboat
lite-efficientnet_lite1_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
12.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
11.0000000 beacon, lighthouse, beacon light, pharos
10.0000000 stupa, tope
9.0000000 liner, ocean liner
8.0000000 catamaran
lite-efficientnet_lite2_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
19.0000000 liner, ocean liner
11.0000000 fireboat
10.0000000 beacon, lighthouse, beacon light, pharos
9.0000000 dock, dockage, docking facility
9.0000000 drilling platform, offshore rig
lite-efficientnet_lite3_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
47.0000000 beacon, lighthouse, beacon light, pharos
33.0000000 liner, ocean liner
15.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
8.0000000 container ship, containership, container vessel
7.0000000 lifeboat
lite-efficientnet_lite4_uint8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
68.0000000 drilling platform, offshore rig
38.0000000 container ship, containership, container vessel
14.0000000 lifeboat
13.0000000 fireboat
12.0000000 dock, dockage, docking facility
efficientnet_lite0_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
53.0000000 lifeboat
34.0000000 catamaran
23.0000000 speedboat
15.0000000 liner, ocean liner
12.0000000 fireboat
efficientnet_lite1_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
12.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
11.0000000 beacon, lighthouse, beacon light, pharos
10.0000000 stupa, tope
9.0000000 liner, ocean liner
8.0000000 catamaran
efficientnet_lite2_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
19.0000000 liner, ocean liner
11.0000000 fireboat
10.0000000 beacon, lighthouse, beacon light, pharos
9.0000000 dock, dockage, docking facility
9.0000000 drilling platform, offshore rig
efficientnet_lite3_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
47.0000000 beacon, lighthouse, beacon light, pharos
33.0000000 liner, ocean liner
15.0000000 breakwater, groin, groyne, mole, bulwark, seawall, jetty
8.0000000 container ship, containership, container vessel
7.0000000 lifeboat
efficientnet_lite4_int8_2.tflite Mean: [127.0, 127.0, 127.0].
Input scale: [128.0].
68.0000000 drilling platform, offshore rig
38.0000000 container ship, containership, container vessel
14.0000000 lifeboat
13.0000000 fireboat
12.0000000 dock, dockage, docking facility

Semantic segmentation

Test image #4

Data source: [PASCAL VOC 2012][PASCAL_VOC_2012]

Image resolution: 500 x 375 

Model Python API
deeplabv3

Color map:

Other tasks

[TBD]