Day 38 - Predict an Image Using NASNetMobile Pre-trained Model for Mobile - IT Consultant - SAP, Artificial Intelligence and Machine Learning

Image Source: Pexels

NASNetMobile architecture is another mobile architecture Convolutional Neural Networks – Layers, Filters, and Architectures,

We have already gone through Convolutional Neural Networks – Layers, Filters, and Architectures, Predict Image Using ResNet50 Pretrained Model, Predict An Image Using InceptionV3 Pretrained Model, Predict an Image Using MobileNetV3 Pretrained Model for Mobile, and Predict An Image Using VGG19 Pretrained Model in the previous blogs. We don’t need to create a new architecture from scratch for our projects; instead, we can use existing Convolutional Neural Networks(CNN) popular architectures like LeNet-5, ResNet, VGG16, VGG19, InceptionV3, MobileNetV3, NASNetMobile, etc. models to check how it predicts.

Today, we will use Convolutional Neural Networks(CNN) NASNetMobile architecture pre-trained model to predict “Flamingo” and check how much accuracy shows. NasNetMobile architecture is specially designed and tuned for Mobile phone CPUs.

We will compare NASNetMobile with ResNet50. VGG16 and VGG19 by Oxford Visual Geometry Group, MobileNetV3, and InceptionV3.

I am working on a few kinds of research on the Enterprise Resource Planning side and using Convolutional Neural Networks(CNN) Image Recognition and computer vision.

I have used Google Colab IDE for this example, and you can download it from the GitHub Repository.

Copy to Clipboard

Copy to Clipboard

array([[[[-0.8039216 , -0.47450978, -0.6313726 ],
         [-0.7647059 , -0.6313726 , -0.69411767],
         [-0.7176471 , -0.6784314 , -0.7254902 ],
         ...,
         [-0.32549018, -0.1372549 , -0.1372549 ],
         [-0.2235294 , -0.12941176, -0.1607843 ],
         [-0.70980394, -0.58431375, -0.60784316]],

        [[-0.6392157 , -0.4588235 , -0.6862745 ],
         [-0.73333335, -0.6627451 , -0.7019608 ],
         [-0.8666667 , -0.7411765 , -0.827451  ],
         ...,
         [-0.372549  , -0.21568626, -0.2862745 ],
         [-0.4823529 , -0.35686272, -0.38039213],
         [-0.70980394, -0.5686275 , -0.5686275 ]],

        [[-0.8745098 , -0.7411765 , -0.81960785],
         [-0.73333335, -0.654902  , -0.7254902 ],
         [-0.92156863, -0.79607844, -0.81960785],
         ...,
         [-0.3490196 , -0.21568626, -0.29411763],
         [-0.6156863 , -0.45098037, -0.49019605],
         [-0.78039217, -0.62352943, -0.5686275 ]],

        ...,

        [[-0.9764706 , -1.        , -0.8980392 ],
         [-0.15294117, -0.1372549 , -0.6313726 ],
         [-1.        , -0.99215686, -1.        ],
         ...,
         [-0.20784312, -0.2862745 , -0.30196077],
         [ 0.12156868,  0.09019613,  0.22352946],
         [-0.08235294, -0.12156862, -0.16862744]],

        [[-0.9843137 , -1.        , -0.88235295],
         [-0.14509803, -0.05882353, -0.654902  ],
         [-0.92156863, -0.9137255 , -0.9529412 ],
         ...,
         [-0.21568626, -0.30196077, -0.3333333 ],
         [-0.11372548, -0.1372549 , -0.20784312],
         [-0.15294117, -0.23137254, -0.3098039 ]],

        [[-0.99215686, -1.        , -0.85882354],
         [ 0.3411765 ,  0.39607847, -0.62352943],
         [-0.9529412 , -0.8352941 , -0.9843137 ],
         ...,
         [ 0.01176476, -0.10588235, -0.14509803],
         [-0.17647058, -0.35686272, -0.40392154],
         [-0.20784312, -0.19999999, -0.26274508]]]], dtype=float32)

Copy to Clipboard

WARNING:tensorflow:5 out of the last 7 calls to <function Model.make_predict_function.<locals>.predict_function at 0x7ff629f43950> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.

array([[1.21793710e-04, 2.30538732e-04, 6.30286086e-05, 4.82737014e-05,
        4.20610049e-05, 8.47176561e-05, 8.36491745e-05, 1.61086151e-04,
        1.89485392e-04, 4.12297377e-04, 2.57266092e-05, 3.27676913e-04,
        2.21578939e-05, 3.75525269e-05, 3.60387712e-05, 8.76136037e-05,
        5.51888479e-05, 1.98003618e-05, 4.66862293e-05, 6.08605988e-05,
        3.95921743e-05, 3.90337809e-05, 3.88692170e-05, 2.56482672e-05,
        3.34303077e-05, 6.34922399e-05, 3.85831045e-05, 1.13853384e-04,
        4.15352770e-05, 4.62669777e-05, 2.62503618e-05, 6.15840472e-05,
        4.72091706e-05, 8.89039075e-05, 9.21786850e-05, 3.70795751e-05,
        5.32145314e-05, 4.88101905e-05, 8.53159290e-05, 5.64221809e-05,
        5.59201326e-05, 6.30385257e-05, 2.94412093e-05, 4.97651963e-05,
        3.59565493e-05, 1.79783441e-04, 7.29523163e-05, 6.18167687e-05,
        8.90196316e-05, 5.43365932e-05, 4.81420757e-05, 7.00715027e-05,
        6.83537655e-05, 8.71957454e-05, 7.15653732e-05, 7.27960069e-05,
        9.67815140e-05, 2.84948255e-05, 5.83787732e-05, 7.29807725e-05,
        1.62609882e-04, 5.85512280e-05, 8.92757089e-05, 6.10842326e-05,
        3.26155314e-05, 1.00788224e-04, 7.87027020e-05, 6.93715483e-05,
        9.43124614e-05, 3.48719586e-05, 6.04818451e-05, 3.93055743e-05,
        4.49765612e-05, 6.90481975e-05, 6.01253378e-05, 1.04397572e-04,
        7.37535665e-05, 4.57502829e-05, 5.04567470e-05, 5.43554052e-05,
        5.02595067e-05, 2.45467472e-05, 1.96995024e-05, 8.32445876e-05,
        3.88380176e-05, 5.23531562e-05, 3.04065543e-05, 6.22747539e-05,
        2.53857317e-04, 3.50628761e-05, 7.16748691e-05, 4.47299208e-05,
        9.32775438e-05, 1.33811307e-04, 5.61124471e-05, 5.08994490e-05,
        2.37266184e-04, 4.76771856e-05, 7.01685931e-05, 3.27750982e-04,
        1.60732528e-03, 3.98107150e-05, 7.16049763e-05, 4.16481598e-05,
        8.45417089e-05, 2.78300722e-05, 4.37346498e-05, 4.43790341e-05,
        1.24818063e-04, 1.14821443e-04, 4.38407988e-05, 9.71839327e-05,
        8.76541380e-05, 5.35162253e-05, 1.50987384e-04, 5.47053933e-05,
        6.25737885e-05, 3.84082814e-05, 3.38475438e-05, 4.32492088e-05,
        5.11421022e-05, 4.06298386e-05, 7.02737379e-05, 4.44807702e-05,
        2.28941171e-05, 2.21165774e-05, 4.56265807e-05, 2.72934674e-04,
        2.38796420e-04, 2.43571034e-04, 9.21408176e-01, 1.33103531e-04,
        1.63078585e-04, 1.60092211e-04, 7.46698352e-04, 1.10401437e-04,
        4.29891115e-05, 5.59826694e-05, 1.57748858e-04, 5.20714748e-05,
        3.63420440e-05, 7.40930191e-05, 6.84581537e-05, 4.93287807e-05,
        2.83639936e-04, 1.10858062e-04, 4.89628765e-05, 6.78677097e-05,
        4.10754983e-05, 8.86712223e-05, 4.28575040e-05, 7.55538640e-05,
        5.69184740e-05, 3.57795325e-05, 6.02691471e-05, 1.16362644e-04,
        3.52672068e-05, 4.56146627e-05, 2.83416721e-05, 6.73628820e-05,
        6.87313295e-05, 4.53942921e-05, 4.41438351e-05, 7.31177352e-05,
        3.89679444e-05, 3.29026989e-05, 3.55880074e-05, 6.59504949e-05,
        4.38883653e-05, 5.44304676e-05, 4.81422139e-05, 5.95234342e-05,
        7.07194049e-05, 1.54373934e-04, 5.14903004e-05, 3.42428357e-05,
        1.05414627e-04, 4.27282612e-05, 3.59829282e-05, 2.71169010e-05,
        4.37168419e-05, 3.96804717e-05, 3.77396827e-05, 1.35275695e-04,
        5.86898168e-05, 1.42651697e-04, 1.45806087e-04, 7.14215203e-05,
        5.64846268e-05, 8.57336127e-05, 7.68700193e-05, 1.68484294e-05,
        8.16335596e-05, 3.46755696e-05, 1.25711871e-04, 2.92424793e-05,
        3.68806723e-05, 7.16897703e-05, 6.95862982e-05, 1.50551859e-04,
        5.73476027e-05, 3.77604920e-05, 4.06738000e-05, 5.92319193e-05,
        8.69702635e-05, 3.96166106e-05, 6.42523082e-05, 4.84041193e-05,
        2.95351947e-05, 6.38981292e-05, 2.70903292e-05, 6.29166025e-05,
        6.07289840e-05, 3.64431435e-05, 4.37703275e-05, 9.15429409e-05,
        4.94793640e-05, 4.76757268e-05, 2.98754803e-05, 5.07064287e-05,
        7.12300316e-05, 4.10262510e-05, 4.73273212e-05, 7.77682071e-05,
        9.38599696e-05, 1.05718573e-04, 4.90241327e-05, 6.23191372e-05,
        5.95261627e-05, 2.94108158e-05, 3.92725269e-05, 5.16896107e-05,
        6.41550141e-05, 8.64266331e-05, 4.23730271e-05, 7.29985986e-05,
        3.25580695e-05, 3.03440975e-05, 4.90234743e-05, 5.85768066e-05,
        6.65075713e-05, 6.91735549e-05, 3.48586254e-05, 6.09782437e-05,
        8.15261883e-05, 1.87724054e-05, 7.32002736e-05, 7.57519156e-05,
        7.58948008e-05, 5.34325882e-05, 1.13268572e-04, 7.69653852e-05,
        6.72926981e-05, 1.06285079e-04, 6.89063017e-05, 9.59318786e-05,
        5.36824555e-05, 2.85062160e-05, 5.71678393e-05, 1.23675709e-04,
        9.05113848e-05, 9.60725956e-05, 4.56200542e-05, 5.90920499e-05,
        5.57357562e-05, 4.33918613e-05, 4.05891333e-05, 3.92677321e-05,
        4.66064230e-05, 4.07473744e-05, 8.51136574e-05, 5.83731489e-05,
        2.94461242e-05, 1.37128591e-04, 1.77056092e-04, 1.42981429e-04,
        4.44958714e-05, 9.02890388e-05, 8.29784258e-05, 8.75943879e-05,
        6.35100441e-05, 5.71655000e-05, 1.05724816e-04, 8.46601470e-05,
        4.91218016e-05, 6.83941325e-05, 3.21443113e-05, 2.14995689e-05,
        3.26536538e-05, 4.31360131e-05, 2.78392308e-05, 1.02458325e-04,
        1.61030548e-04, 2.00243649e-05, 5.28714081e-05, 1.15707880e-04,
        1.73963170e-04, 8.52062440e-05, 3.57424615e-05, 3.28403803e-05,
        4.89933300e-05, 1.16137584e-04, 4.64500517e-05, 4.83059375e-05,
        7.41799086e-05, 3.34010547e-05, 3.13144010e-05, 7.91966886e-05,
        6.35733013e-05, 7.65298973e-05, 7.90111881e-05, 4.47597922e-05,
        4.68139187e-05, 4.43339850e-05, 5.74560509e-05, 5.03646843e-05,
        6.97031937e-05, 5.96705286e-05, 5.51667472e-05, 5.53932805e-05,
        1.03040191e-04, 5.01071263e-05, 5.56837986e-05, 5.65387381e-05,
        3.93882001e-05, 3.90343193e-04, 6.69653746e-05, 7.01817844e-05,
        4.71171006e-05, 5.57993117e-05, 4.29213542e-05, 4.75165398e-05,
        9.87356398e-05, 7.34082496e-05, 8.89188304e-05, 6.31435760e-05,
        1.87113837e-05, 2.76969968e-05, 3.27602684e-05, 7.75635563e-05,
        5.44139111e-05, 5.89005358e-05, 9.48458910e-05, 4.18436393e-05,
        5.91284697e-05, 3.46817542e-05, 4.01642610e-05, 2.47725093e-05,
        4.35311267e-05, 6.03625631e-05, 5.18587221e-05, 1.08721651e-04,
        1.69421008e-04, 1.67561579e-04, 1.90163715e-04, 9.67214437e-05,
        5.11549806e-05, 2.66529914e-05, 4.50392290e-05, 6.66426131e-05,
        3.67470493e-05, 8.64572139e-05, 6.69238725e-05, 2.42802107e-05,
        4.44393008e-05, 2.71967219e-05, 5.55985789e-05, 3.11383446e-05,
        6.36613986e-05, 3.41762861e-05, 3.70887028e-05, 5.51687990e-05,
        2.99595067e-05, 3.31045449e-05, 3.08721064e-05, 4.49296604e-05,
        8.36743056e-05, 2.91866436e-05, 8.02523718e-05, 6.36745172e-05,
        4.40566509e-05, 1.21843550e-04, 1.09343244e-04, 6.14165983e-05,
        7.40574105e-05, 6.96752832e-05, 4.64912700e-05, 4.82929972e-05,
        7.63081916e-05, 3.97223084e-05, 7.12885449e-05, 3.93879745e-05,
        9.49081441e-05, 1.32649104e-04, 6.23912711e-05, 2.95582140e-05,
        8.50482611e-05, 2.47614553e-05, 3.60070590e-05, 3.35992954e-05,
        1.10537127e-04, 1.15750372e-04, 1.28955711e-04, 3.83594852e-05,
        4.22113735e-05, 5.86403003e-05, 4.96849570e-05, 5.88783514e-05,
        7.53557688e-05, 6.56286284e-05, 4.60475821e-05, 2.26614182e-04,
        3.56473829e-05, 5.86839960e-05, 5.44542454e-05, 4.07495900e-05,
        3.27281014e-05, 1.24427519e-04, 4.13398993e-05, 8.38494889e-05,
        8.28710326e-05, 9.20265593e-05, 6.48657078e-05, 1.11422036e-04,
        7.75587469e-05, 1.51585118e-04, 5.38181921e-05, 2.95040318e-05,
        9.75185394e-05, 1.16208221e-05, 8.70882868e-05, 2.86746363e-05,
        3.73737093e-05, 1.81348398e-04, 3.53900505e-05, 4.99458220e-05,
        1.28456275e-04, 4.95938293e-05, 6.31086587e-05, 8.90544397e-05,
        3.30405164e-05, 6.36481636e-05, 5.09044003e-05, 7.83572250e-05,
        7.17223156e-05, 6.40372818e-05, 1.62278429e-05, 6.09230228e-05,
        1.11270827e-04, 6.78717188e-05, 6.80002704e-05, 7.64306242e-05,
        5.89203664e-05, 3.18137791e-05, 4.50705957e-05, 5.34518040e-05,
        6.32403026e-05, 5.28225719e-05, 5.74096048e-05, 8.33616068e-05,
        3.03636098e-05, 7.82952266e-05, 5.25706419e-05, 6.04312190e-05,
        6.43748499e-05, 7.42658449e-05, 4.36635528e-05, 4.21945915e-05,
        1.23184844e-04, 3.53444484e-05, 8.22800139e-05, 4.32074849e-05,
        7.76130619e-05, 5.45133225e-05, 1.09970810e-04, 5.91439202e-05,
        1.71437059e-04, 1.21596182e-04, 8.08507175e-05, 8.22865259e-05,
        4.21005352e-05, 9.80954501e-05, 1.62707613e-04, 6.92877729e-05,
        8.04701049e-05, 9.45482243e-05, 6.15466561e-05, 3.65266496e-05,
        7.28294835e-05, 2.90682565e-05, 4.14051210e-05, 4.99870439e-05,
        4.26048864e-05, 3.85379826e-05, 7.39197494e-05, 6.85218329e-05,
        1.15934301e-04, 3.61781349e-05, 5.01418290e-05, 6.81051461e-05,
        7.45442885e-05, 3.69033987e-05, 4.66439051e-05, 8.20688147e-05,
        3.29846989e-05, 1.00391931e-04, 4.26641272e-05, 4.92993386e-05,
        7.06957435e-05, 3.98955053e-05, 2.91262240e-05, 4.22305820e-05,
        9.05363340e-05, 9.38253361e-05, 3.77189936e-05, 7.90536942e-05,
        3.76324715e-05, 6.26419715e-05, 5.84986483e-05, 3.55017037e-05,
        2.04578264e-05, 7.45182697e-05, 1.73077016e-04, 1.28616113e-04,
        5.05820608e-05, 4.46358717e-05, 3.87631226e-05, 5.35722938e-05,
        3.26210356e-05, 3.54288823e-05, 1.25276842e-04, 5.65664595e-05,
        4.59854891e-05, 6.80491794e-05, 3.67161520e-05, 4.80760100e-05,
        3.18266466e-05, 7.43591081e-05, 1.30322092e-04, 5.25192772e-05,
        4.97133042e-05, 6.94021874e-05, 4.96047578e-05, 1.89373197e-04,
        9.22157051e-05, 7.78252579e-05, 8.44128081e-05, 5.65639239e-05,
        5.13980194e-05, 5.98198094e-05, 1.21538214e-04, 7.15427232e-05,
        2.79169588e-04, 5.86048554e-05, 2.44000094e-05, 8.23741502e-05,
        4.57863789e-05, 1.52118344e-04, 1.57216073e-05, 2.82163219e-05,
        1.03153063e-04, 5.51930571e-05, 1.03498642e-04, 1.57481227e-05,
        1.16114112e-04, 3.16271326e-05, 5.67815223e-05, 7.42036864e-05,
        4.82090181e-05, 5.06788238e-05, 8.16392348e-05, 7.71172490e-05,
        1.14784329e-04, 5.80563647e-05, 1.74574772e-04, 3.41234583e-04,
        4.70934247e-05, 7.92368810e-05, 7.81416311e-05, 3.99225646e-05,
        1.54557027e-04, 8.44682072e-05, 4.01273901e-05, 4.25937142e-05,
        1.32306974e-04, 1.19133809e-04, 4.35393049e-05, 4.76286477e-05,
        2.77895215e-05, 5.28169330e-05, 4.17142255e-05, 4.91929168e-05,
        4.96379289e-05, 1.13570350e-04, 6.28431444e-05, 4.17814736e-05,
        4.84678167e-05, 6.97651776e-05, 3.78644181e-05, 5.93287587e-05,
        1.54285328e-04, 1.70170097e-04, 6.39034988e-05, 3.89960478e-05,
        2.15125729e-05, 9.86412342e-05, 1.00999132e-04, 2.03007818e-04,
        9.75588191e-05, 1.08158136e-04, 6.94917253e-05, 3.79035118e-05,
        6.20648061e-05, 7.34572750e-05, 8.17423133e-05, 5.45977164e-05,
        8.77653001e-05, 3.18906314e-05, 8.46049443e-05, 1.08249558e-04,
        8.48228301e-05, 7.52192063e-05, 1.06544278e-04, 2.91484248e-05,
        4.78487273e-05, 4.35121583e-05, 5.31247824e-05, 1.77913069e-04,
        1.87834721e-05, 7.84433578e-05, 4.45825681e-05, 4.86198041e-05,
        1.27818988e-04, 7.60444818e-05, 1.03275117e-04, 7.90781269e-05,
        1.13695729e-04, 1.09041881e-04, 1.16381067e-04, 8.57117848e-05,
        6.81966558e-05, 4.92042273e-05, 4.05641731e-05, 7.85761076e-05,
        1.04584215e-04, 8.98463404e-05, 3.04538436e-04, 8.67775598e-05,
        6.50985967e-05, 8.35773317e-05, 8.65188267e-05, 7.54629218e-05,
        5.33078928e-05, 3.34744618e-05, 1.14841489e-04, 2.40889276e-04,
        7.70137704e-05, 3.55325938e-05, 4.30983055e-05, 1.85193960e-04,
        5.25542018e-05, 5.44270915e-05, 6.53083407e-05, 2.97757560e-05,
        9.48533998e-05, 4.65824705e-05, 3.99283927e-05, 9.10339877e-05,
        9.19227023e-05, 7.89233745e-05, 3.23130807e-05, 8.31203579e-05,
        1.41408134e-04, 3.48177273e-05, 9.48050147e-05, 8.95360135e-05,
        5.11045146e-05, 5.38867025e-05, 5.11304970e-05, 5.95416022e-05,
        6.95509298e-05, 5.77144347e-05, 9.42625556e-05, 5.22160844e-05,
        4.82682262e-05, 7.71998093e-05, 4.90186139e-05, 2.03282434e-05,
        8.22356160e-05, 1.03202357e-04, 6.05622445e-05, 8.68807401e-05,
        5.54777616e-05, 5.77231294e-05, 5.12543556e-05, 1.05483516e-04,
        6.54780961e-05, 7.71800114e-05, 1.29675929e-04, 2.03437176e-05,
        2.92419500e-05, 1.03000100e-04, 9.15270502e-05, 7.99245536e-05,
        4.17305018e-05, 7.84759104e-05, 5.58281608e-05, 6.70783775e-05,
        6.13433731e-05, 5.81007334e-05, 8.50113574e-05, 1.77333743e-04,
        5.11725957e-05, 5.46535084e-05, 3.27845519e-05, 3.34879369e-05,
        3.90781788e-05, 4.39748401e-05, 2.72682719e-05, 4.95783643e-05,
        3.85996682e-05, 1.69473686e-04, 6.85664199e-05, 1.46310311e-04,
        3.75687887e-05, 8.48958298e-05, 4.35228249e-05, 1.72308079e-04,
        7.42351112e-05, 5.27094016e-05, 7.11898319e-05, 7.42416960e-05,
        1.13906710e-04, 5.75351187e-05, 7.40933756e-05, 6.16657417e-05,
        2.19319318e-05, 6.65884945e-05, 4.92624931e-05, 8.26630421e-05,
        8.76237173e-05, 2.86261929e-05, 4.45669248e-05, 3.89283523e-05,
        2.92264758e-05, 8.72059754e-05, 5.29292774e-05, 6.41935694e-05,
        4.77578214e-05, 5.22459704e-05, 6.62524544e-05, 1.03601145e-04,
        2.25650074e-05, 2.26018146e-05, 8.44528258e-05, 6.85122286e-05,
        8.01505666e-05, 7.03599726e-05, 2.53486578e-05, 7.84032673e-05,
        4.12634836e-05, 6.19422281e-05, 8.94439290e-05, 5.91914322e-05,
        9.46682339e-05, 5.24245515e-05, 4.35779803e-05, 5.63574831e-05,
        5.06025171e-05, 5.06543729e-05, 7.40384930e-05, 6.24311506e-05,
        1.24444734e-04, 3.09071620e-05, 1.03306636e-04, 1.18862132e-04,
        3.08157469e-05, 1.88354552e-05, 4.36571790e-05, 1.16637515e-04,
        3.02005865e-05, 3.60636623e-05, 4.87157849e-05, 4.43759454e-05,
        9.04203625e-05, 2.86870800e-05, 8.69681899e-05, 8.93630131e-05,
        8.34319144e-05, 2.34747495e-05, 1.83572611e-05, 5.36276020e-05,
        1.57879796e-04, 1.11131376e-04, 2.20517941e-05, 1.03594328e-04,
        3.68479778e-05, 6.61387603e-05, 1.90036517e-05, 5.35503787e-05,
        9.39371603e-05, 1.01975020e-04, 9.03030814e-05, 8.93407778e-05,
        4.47241182e-05, 5.42029593e-05, 3.84062296e-05, 4.32194392e-05,
        8.10349011e-05, 3.89106062e-05, 7.51277257e-05, 3.02784520e-05,
        7.62194686e-05, 8.43383750e-05, 1.09846915e-04, 1.04756007e-04,
        3.19619430e-05, 1.09678076e-04, 6.75600677e-05, 8.98059880e-05,
        4.73265536e-05, 3.71329043e-05, 1.29487846e-04, 1.20059718e-04,
        3.87860528e-05, 6.79512596e-05, 6.98073709e-05, 1.19120858e-04,
        9.73908391e-05, 1.27900203e-04, 6.59507496e-05, 7.40233809e-05,
        5.79533626e-05, 2.64304745e-05, 9.51304246e-05, 7.54430584e-05,
        1.27945837e-04, 1.56053982e-04, 8.69696014e-05, 4.84348275e-05,
        6.00786843e-05, 4.01612378e-05, 1.76396425e-04, 2.70412740e-04,
        1.75845344e-04, 3.72120994e-05, 9.58873352e-05, 3.94371418e-05,
        2.12099447e-04, 4.31149929e-05, 6.66653068e-05, 9.98681644e-05,
        2.59322424e-05, 8.43672533e-05, 5.84003792e-05, 4.73353575e-05,
        3.20820182e-05, 6.75136253e-05, 3.58068392e-05, 8.44560491e-05,
        5.40463006e-05, 2.36451961e-05, 8.59682623e-05, 3.65018568e-05,
        6.04288616e-05, 6.62759558e-05, 3.83024599e-05, 4.43793324e-05,
        1.74934918e-04, 6.97962460e-05, 5.77297942e-05, 8.89617513e-05,
        4.25819781e-05, 2.10315268e-03, 5.52994861e-05, 7.33999914e-05,
        6.73497125e-05, 6.11602445e-05, 7.42880875e-05, 1.89370679e-04,
        1.22357422e-04, 3.36009907e-05, 9.95708833e-05, 7.34099376e-05,
        1.42881938e-04, 1.20329998e-04, 5.60448461e-05, 3.32798481e-05,
        4.27627056e-05, 7.34723362e-05, 6.04436100e-05, 2.08957179e-04,
        5.49087163e-05, 5.36101143e-05, 1.25538878e-04, 2.43721261e-05,
        9.27985820e-05, 8.41062501e-05, 3.38656901e-05, 7.36557922e-05,
        6.40066937e-05, 5.04898162e-05, 1.60997239e-04, 1.95542816e-04,
        8.96995116e-05, 2.55041141e-05, 9.82514248e-05, 6.54081887e-05,
        4.65630619e-05, 7.89791666e-05, 4.39192227e-05, 8.58123021e-05,
        2.84988746e-05, 4.22385165e-05, 6.65911575e-05, 7.34272253e-05,
        1.05641564e-04, 5.79685075e-05, 3.33574753e-05, 5.73876459e-05,
        2.18996011e-05, 7.62646887e-05, 5.64350375e-05, 2.14086540e-04,
        8.27266049e-05, 7.23611302e-05, 4.56096604e-05, 7.20216412e-05,
        6.33365707e-05, 9.01013555e-05, 3.49368711e-05, 9.92923815e-05,
        1.31827619e-04, 8.30210120e-05, 5.55554871e-05, 7.95068045e-05,
        1.65744364e-04, 3.89299457e-05, 4.31890294e-05, 9.73228671e-05,
        1.57541392e-04, 6.01933862e-05, 1.28791144e-04, 6.16652687e-05,
        3.81572099e-05, 4.91479514e-05, 5.14825442e-05, 1.41133569e-04,
        1.39839336e-04, 5.06659708e-05, 4.16619092e-04, 8.59910215e-04,
        1.10956847e-04, 3.25170768e-05, 7.24891797e-05, 1.93364976e-04,
        2.35516436e-05, 1.69612424e-04, 3.67210196e-05, 8.81206579e-05,
        8.95782869e-05, 1.89062659e-04, 7.17260118e-05, 2.61385248e-05,
        8.99075385e-05, 2.75282000e-05, 9.90394983e-05, 1.00887759e-04,
        8.11761420e-05, 1.40250966e-04, 8.42028749e-05, 1.30797736e-04,
        6.67192435e-05, 6.98589793e-05, 3.47328605e-05, 8.07704128e-05,
        3.14998942e-05, 1.33422393e-04, 1.03749153e-04, 8.17835535e-05,
        1.40662582e-04, 3.39732469e-05, 4.65416160e-05, 6.75572956e-05,
        1.18283831e-04, 7.24761194e-05, 4.15212737e-04, 7.45934303e-05,
        3.71009410e-05, 1.72309548e-04, 7.90638005e-05, 6.93331240e-05,
        7.57084345e-05, 3.60434424e-05, 1.23334015e-04, 4.51118358e-05,
        9.46750879e-05, 1.06769476e-04, 7.93746804e-05, 4.27499035e-05]],
      dtype=float32)

Copy to Clipboard

[[('n02007558', 'flamingo', 0.9214082),
  ('n04486054', 'triumphal_arch', 0.0021031527),
  ('n01860187', 'black_swan', 0.0016073253),
  ('n07749582', 'lemon', 0.0008599102),
  ('n02012849', 'crane', 0.00074669835)]]

Hurray! it predicted as “Flamingo” with 0.92 prediction accuracy.

Let’s check the prediction accuracy on VGG19, VGG16, InceptionV3, MobileNetV3, and ResNet50.

VGG16:

[[('n02012849', 'crane', 0.7303488),
  ('n02009912', 'American_egret', 0.15676753),
  ('n02007558', 'flamingo', 0.067069136),
  ('n02009229', 'little_blue_heron', 0.03742734),
  ('n03388043', 'fountain', 0.002747606)]]

VGG19:

[[('n02009912', 'American_egret', 0.49541578),
  ('n02012849', 'crane', 0.37884706),
  ('n02009229', 'little_blue_heron', 0.11860012),
  ('n02007558', 'flamingo', 0.0025609955),
  ('n02006656', 'spoonbill', 0.0023633465)]]

InceptionV3:

[('n02007558', 'flamingo', 0.94305456),
  ('n02012849', 'crane', 0.0009208125),
  ('n04550184', 'wardrobe', 0.0006756062),
  ('n02009912', 'American_egret', 0.00046367754),
  ('n01860187', 'black_swan', 0.00044391607)]]

MobileNetV3:

[[('n02007558', 'flamingo', 0.62121254),
  ('n02012849', 'crane', 0.030542558),
  ('n02002724', 'black_stork', 0.009830186),
  ('n13040303', 'stinkhorn', 0.005730417),
  ('n03532672', 'hook', 0.0049227206)]]

ResNet50:

[[('n03388043', 'fountain', 0.54863626),
  ('n02007558', 'flamingo', 0.055300727),
  ('n01806143', 'peacock', 0.04455783),
  ('n04209133', 'shower_cap', 0.0367395),
  ('n04507155', 'umbrella', 0.033092268)]]

In this example, we could see NASNetMobile InceptionV3, and MobileNetV3 
pretrained models predicts as "Flamingo".

Day 38 – Predict an Image Using NASNetMobile Pre-trained Model for Mobile

Day 38 – Predict an Image Using NASNetMobile Pre-trained Model for Mobile

Further Reading

Posts on Artificial Intelligence, Deep Learning, Machine Learning, and Design Thinking articles:

Posts on SAP:

About the Author: Gopi Chandrakesan

Leave A Comment Cancel reply

Day 38 – Predict an Image Using NASNetMobile Pre-trained Model for Mobile

Further Reading

Posts on Artificial Intelligence, Deep Learning, Machine Learning, and Design Thinking articles:

Posts on SAP:

Share This Blog, Choose Your Platform!

About the Author: Gopi Chandrakesan

Related Posts

What are Simulated Annealing (SA), Local Beam Search, Genetic, and Hill-climbing Search Algorithms?

Types of Local Search Algorithms

How do we measure the performance of Problem-Solving Algorithms?

Different types of Real-world problems in Problem Solving approach

Leave A Comment Cancel reply