Example #1
0
# dropout
if use.drop:
    drop.p_vid = shared(float32(drop.p_vid_val))
    drop.p_hidden = shared(float32(drop.p_hidden_val))
    drop.p_vid.set_value(float32(0.))  # dont use dropout when testing
    drop.p_hidden.set_value(float32(0.))  # dont use dropout when testing
    vid_ = DropoutLayer(vid_, rng=tr.rng, p=drop.p_vid).output

# MLP
# ------------------------------------------------------------------------------
# fusion
if net.fusion == "early":
    out = vid_
    # hidden layer
    Wh, bh = load_params(use)  # This is test, wudi added this!
    layers.append(
        HiddenLayer(out,
                    W=Wh,
                    b=bh,
                    n_in=n_in_MLP,
                    n_out=net.hidden,
                    rng=tr.rng,
                    W_scale=net.W_scale[-2],
                    b_scale=net.b_scale[-2],
                    activation=relu))
    out = layers[-1].output

if tr.inspect:
    insp = T.stack(insp[0], insp[1], insp[2], insp[3], insp[4], insp[5],
                   T.mean(out))
Example #2
0
    def __init__(self,
                 x,
                 use,
                 lr,
                 batch,
                 net,
                 reg,
                 drop,
                 mom,
                 tr,
                 res_dir,
                 load_path=""):

        self.out = []
        self.layers = []
        self.insp_mean = []
        self.insp_std = []

        for c in (use, lr, batch, net, reg, drop, mom, tr):
            write(c.__name__ + ":", res_dir)
            _s = c.__dict__
            del _s['__module__'], _s['__doc__']
            for key in _s.keys():
                val = str(_s[key])
                if val.startswith("<static"):
                    val = str(_s[key].__func__.__name__)
                if val.startswith("<Cuda"): continue
                if val.startswith("<Tensor"): continue
                write("  " + key + ": " + val, res_dir)

        ####################################################################
        ####################################################################
        print "\n%s\n\tbuilding\n%s" % (('-' * 30, ) * 2)
        ####################################################################
        ####################################################################
        # ConvNet
        # ------------------------------------------------------------------------------
        # calculate resulting video shapes for all stages
        print net.n_stages
        conv_shapes = []
        for i in xrange(net.n_stages):
            k, p, v = array(net.kernels[i]), array(net.pools[i]), array(
                tr.video_shapes[i])
            conv_s = tuple(v - k + 1)
            conv_shapes.append(conv_s)
            tr.video_shapes.append(tuple((v - k + 1) / p))
            print "stage", i
            if use.depth and i == 0:
                print "  conv", tr.video_shapes[
                    i], "x 2 ->", conv_s  #for body and hand
            else:
                print "  conv", tr.video_shapes[i], "->", conv_s
            print "  pool", conv_s, "->", tr.video_shapes[i +
                                                          1], "x", net.maps[i +
                                                                            1]

        # number of inputs for MLP = (# maps last stage)*(# convnets)*(resulting video shape) + trajectory size
        n_in_MLP = net.maps[-1] * net.n_convnets * prod(tr.video_shapes[-1])
        print 'debug1'
        if use.depth:
            if net.n_convnets == 2:
                out = [x[:, :, 0, :, :, :],
                       x[:, :, 1, :, :, :]]  # 2 nets: body and hand

        # build 3D ConvNet

        for stage in xrange(net.n_stages):
            for i in xrange(len(out)):  # for body and hand
                # normalization
                if use.norm and stage == 0:
                    gray_norm = NormLayer(out[i][:, 0:1],
                                          method="lcn",
                                          use_divisor=use.norm_div).output
                    gray_norm = std_norm(gray_norm, axis=[-3, -2, -1])
                    depth_norm = var_norm(out[i][:, 1:])
                    out[i] = T.concatenate([gray_norm, depth_norm], axis=1)
                elif use.norm:
                    out[i] = NormLayer(out[i],
                                       method="lcn",
                                       use_divisor=use.norm_div).output
                    out[i] = std_norm(out[i], axis=[-3, -2, -1])
                # convolutions
                out[i] *= net.scaler[stage][i]
                print 'debug2'
                self.layers.append(
                    ConvLayer(
                        out[i],
                        **conv_args(stage, i, batch, net, use, tr.rng,
                                    tr.video_shapes, load_path)))
                out[i] = self.layers[-1].output
                out[i] = PoolLayer(out[i],
                                   net.pools[stage],
                                   method=net.pool_method).output
                if tr.inspect:
                    self.insp_mean.append(T.mean(out[i]))
                    self.insp_std.append(T.std(out[i]))
        print 'debug2'
        # flatten all convnets outputs
        for i in xrange(len(out)):
            out[i] = std_norm(out[i], axis=[-3, -2, -1])
        out = [out[i].flatten(2) for i in range(len(out))]
        vid_ = T.concatenate(out, axis=1)
        print 'debug3'
        # dropout
        if use.drop:
            drop.p_vid = shared(float32(drop.p_vid_val))
            drop.p_hidden = shared(float32(drop.p_hidden_val))
            vid_ = DropoutLayer(vid_, rng=tr.rng, p=drop.p_hidden).output

        #maxout
        if use.maxout:
            vid_ = maxout(vid_, (batch.micro, n_in_MLP))
            net.activation = lin
            n_in_MLP /= 2
            # net.hidden *= 2

        # MLP
        # ------------------------------------------------------------------------------
        # fusion
        if net.fusion == "early":
            out = vid_
            # hidden layer
            if use.load:
                W, b = load_params(use, load_path)
                self.layers.append(
                    HiddenLayer(out,
                                n_in=n_in_MLP,
                                n_out=net.hidden_vid,
                                rng=tr.rng,
                                W=W,
                                b=b,
                                W_scale=net.W_scale[-2],
                                b_scale=net.b_scale[-2],
                                activation=net.activation))
            else:
                self.layers.append(
                    HiddenLayer(out,
                                n_in=n_in_MLP,
                                n_out=net.hidden_vid,
                                rng=tr.rng,
                                W_scale=net.W_scale[-2],
                                b_scale=net.b_scale[-2],
                                activation=net.activation))
            out = self.layers[-1].output

        #if tr.inspect:
        #self.insp_mean = T.stack(self.insp_mean)
        #self.insp_std = T.stack(self.insp_std)
        #self.insp = T.stack(self.insp[0],self.insp[1],self.insp[2],self.insp[3],self.insp[4],self.insp[5], T.mean(out))
        #else: self.insp =  T.stack(0,0)
        # out = normalize(out)
        if use.drop:
            out = DropoutLayer(out, rng=tr.rng, p=drop.p_hidden).output
        #maxout
        if use.maxout:
            out = maxout(out, (batch.micro, net.hidden))
            net.hidden /= 2
        print 'debug3'

        # now assembly all the output
        self.out = out
        self.n_in_MLP = n_in_MLP
def build():
    use.load = True  # we load the CNN parameteres here
    x = ndtensor(len(tr.in_shape))(name='x')  # video input
    x_ = _shared(empty(tr.in_shape))

    conv_shapes = []
    for i in xrange(net.n_stages):
        k, p, v = array(net.kernels[i]), array(net.pools[i]), array(
            tr.video_shapes[i])
        conv_s = tuple(v - k + 1)
        conv_shapes.append(conv_s)
        tr.video_shapes.append(tuple((v - k + 1) / p))
        print "stage", i
        print "  conv", tr.video_shapes[i], "->", conv_s
        print "  pool", conv_s, "->", tr.video_shapes[i + 1], "x", net.maps[i +
                                                                            1]

    # number of inputs for MLP = (# maps last stage)*(# convnets)*(resulting video shape) + trajectory size
    n_in_MLP = net.maps[-1] * net.n_convnets * prod(tr.video_shapes[-1])
    print 'MLP:', n_in_MLP, "->", net.hidden, "->", net.n_class, ""

    if use.depth:
        if net.n_convnets == 2:
            out = [x[:, :, 0, :, :, :], x[:, :,
                                          1, :, :, :]]  # 2 nets: body and hand

    # build 3D ConvNet
    layers = []  # all architecture layers
    insp = []
    for stage in xrange(net.n_stages):
        for i in xrange(len(out)):  # for body and hand
            # normalization
            if use.norm and stage == 0:
                gray_norm = NormLayer(out[i][:, 0:1],
                                      method="lcn",
                                      use_divisor=use.norm_div).output
                gray_norm = std_norm(gray_norm, axis=[-3, -2, -1])
                depth_norm = var_norm(out[i][:, 1:])
                out[i] = T.concatenate([gray_norm, depth_norm], axis=1)
            elif use.norm:
                out[i] = NormLayer(out[i],
                                   method="lcn",
                                   use_divisor=use.norm_div).output
                out[i] = std_norm(out[i], axis=[-3, -2, -1])
            # convolutions
            out[i] *= net.scaler[stage][i]
            layers.append(
                ConvLayer(
                    out[i],
                    **conv_args(stage, i, batch, net, use, tr.rng,
                                tr.video_shapes)))
            out[i] = layers[-1].output
            out[i] = PoolLayer(out[i],
                               net.pools[stage],
                               method=net.pool_method).output
            if tr.inspect: insp.append(T.mean(out[i]))

    # flatten all convnets outputs
    for i in xrange(len(out)):
        out[i] = std_norm(out[i], axis=[-3, -2, -1])
    out = [out[i].flatten(2) for i in range(len(out))]
    vid_ = T.concatenate(out, axis=1)

    # dropout
    if use.drop:
        drop.p_vid = shared(float32(drop.p_vid_val))
        drop.p_hidden = shared(float32(drop.p_hidden_val))
        drop.p_vid.set_value(float32(0.))  # dont use dropout when testing
        drop.p_hidden.set_value(float32(0.))  # dont use dropout when testing
        vid_ = DropoutLayer(vid_, rng=tr.rng, p=drop.p_vid).output

    # MLP
    # ------------------------------------------------------------------------------
    # fusion
    if net.fusion == "early":
        out = vid_
        # hidden layer
        Wh, bh = load_params(use)  # This is test, wudi added this!
        layers.append(
            HiddenLayer(out,
                        W=Wh,
                        b=bh,
                        n_in=n_in_MLP,
                        n_out=net.hidden,
                        rng=tr.rng,
                        W_scale=net.W_scale[-2],
                        b_scale=net.b_scale[-2],
                        activation=relu))
        out = layers[-1].output

    if tr.inspect:
        insp = T.stack(insp[0], insp[1], insp[2], insp[3], insp[4], insp[5],
                       T.mean(out))
    else:
        insp = T.stack(0, 0)

    if use.drop: out = DropoutLayer(out, rng=tr.rng, p=drop.p_hidden).output
    #maxout
    # softmax layer
    Ws, bs = load_params(use)  # This is test, wudi added this!
    layers.append(
        LogRegr(out,
                W=Ws,
                b=bs,
                rng=tr.rng,
                activation=lin,
                n_in=net.hidden,
                W_scale=net.W_scale[-1],
                b_scale=net.b_scale[-1],
                n_out=net.n_class))
    """
    layers[-1] : softmax layer
    layers[-2] : hidden layer (video if late fusion)
    layers[-3] : hidden layer (trajectory, only if late fusion)
    """
    # prediction
    y_pred = layers[-1].y_pred
    p_y_given_x = layers[-1].p_y_given_x
    ####################################################################
    ####################################################################
    print "\n%s\n\tcompiling\n%s" % (('-' * 30, ) * 2)
    ####################################################################
    ####################################################################
    # compile functions
    # ------------------------------------------------------------------------------
    print 'compiling test_model'

    eval_model = function([], [y_pred, p_y_given_x],
                          givens={x: x_},
                          on_unused_input='ignore')

    return eval_model, x_
def build():
    use.load = True  # we load the CNN parameteres here
    x = ndtensor(len(tr.in_shape))(name = 'x') # video input
    x_ = _shared(empty(tr.in_shape))

    conv_shapes = []
    for i in xrange(net.n_stages):
        k,p,v = array(net.kernels[i]), array(net.pools[i]), array(tr.video_shapes[i])
        conv_s = tuple(v-k+1)
        conv_shapes.append(conv_s)
        tr.video_shapes.append(tuple((v-k+1)/p))
        print "stage", i
        print "  conv",tr.video_shapes[i],"->",conv_s
        print "  pool",conv_s,"->",tr.video_shapes[i+1],"x",net.maps[i+1]

    # number of inputs for MLP = (# maps last stage)*(# convnets)*(resulting video shape) + trajectory size
    n_in_MLP = net.maps[-1]*net.n_convnets*prod(tr.video_shapes[-1]) 
    print 'MLP:', n_in_MLP, "->", net.hidden, "->", net.n_class, ""

    if use.depth:
        if net.n_convnets==2: 
            out = [x[:,:,0,:,:,:], x[:,:,1,:,:,:]] # 2 nets: body and hand

    # build 3D ConvNet
    layers = [] # all architecture layers
    insp = []
    for stage in xrange(net.n_stages):
        for i in xrange(len(out)): # for body and hand
            # normalization
            if use.norm and stage==0: 
                gray_norm = NormLayer(out[i][:,0:1], method="lcn",
                    use_divisor=use.norm_div).output
                gray_norm = std_norm(gray_norm,axis=[-3,-2,-1])
                depth_norm = var_norm(out[i][:,1:])
                out[i]  = T.concatenate([gray_norm,depth_norm],axis=1)
            elif use.norm:
                out[i] = NormLayer(out[i], method="lcn",use_divisor=use.norm_div).output
                out[i] = std_norm(out[i],axis=[-3,-2,-1])
            # convolutions  
            out[i] *= net.scaler[stage][i]
            layers.append(ConvLayer(out[i], **conv_args(stage, i, batch, net, use, tr.rng, tr.video_shapes)))
            out[i] = layers[-1].output
            out[i] = PoolLayer(out[i], net.pools[stage], method=net.pool_method).output
            if tr.inspect: insp.append(T.mean(out[i]))

    # flatten all convnets outputs
    for i in xrange(len(out)): out[i] = std_norm(out[i],axis=[-3,-2,-1])
    out = [out[i].flatten(2) for i in range(len(out))]
    vid_ = T.concatenate(out, axis=1)

    # dropout
    if use.drop: 
        drop.p_vid = shared(float32(drop.p_vid_val) )
        drop.p_hidden = shared(float32(drop.p_hidden_val))
        drop.p_vid.set_value(float32(0.))  # dont use dropout when testing
        drop.p_hidden.set_value(float32(0.))  # dont use dropout when testing
        vid_ = DropoutLayer(vid_, rng=tr.rng, p=drop.p_vid).output

    # MLP
    # ------------------------------------------------------------------------------
    # fusion
    if net.fusion == "early":
        out = vid_
        # hidden layer
        Wh, bh = load_params(use)  # This is test, wudi added this!
        layers.append(HiddenLayer(out, W = Wh, b =bh, n_in=n_in_MLP, n_out=net.hidden, rng=tr.rng, 
            W_scale=net.W_scale[-2], b_scale=net.b_scale[-2], activation=relu))
        out = layers[-1].output

    if tr.inspect: insp = T.stack(insp[0],insp[1],insp[2],insp[3],insp[4],insp[5], T.mean(out))
    else: insp =  T.stack(0,0)

    if use.drop: out = DropoutLayer(out, rng=tr.rng, p=drop.p_hidden).output
    #maxout
    # softmax layer
    Ws, bs = load_params(use) # This is test, wudi added this!
    layers.append(LogRegr(out, W = Ws, b = bs, rng=tr.rng, activation=lin, n_in=net.hidden, 
        W_scale=net.W_scale[-1], b_scale=net.b_scale[-1], n_out=net.n_class))
    """
    layers[-1] : softmax layer
    layers[-2] : hidden layer (video if late fusion)
    layers[-3] : hidden layer (trajectory, only if late fusion)
    """
    # prediction
    y_pred = layers[-1].y_pred
    p_y_given_x = layers[-1].p_y_given_x
    ####################################################################
    ####################################################################
    print "\n%s\n\tcompiling\n%s"%(('-'*30,)*2)
    ####################################################################
    #################################################################### 
    # compile functions
    # ------------------------------------------------------------------------------
    print 'compiling test_model'

    eval_model = function([], [y_pred, p_y_given_x], 
        givens={x:x_},
        on_unused_input='ignore')

    return eval_model, x_
    def __init__(self, x, use, lr, batch, net, reg, drop, mom, tr, res_dir, load_path=""):

        self.out = []
        self.layers = []
        self.insp_mean = []
        self.insp_std = []

        for c in (use, lr, batch, net, reg, drop, mom, tr):
            write(c.__name__+":", res_dir)
            _s = c.__dict__
            del _s['__module__'], _s['__doc__']
            for key in _s.keys(): 
                val = str(_s[key])
                if val.startswith("<static"): val = str(_s[key].__func__.__name__)
                if val.startswith("<Cuda"): continue
                if val.startswith("<Tensor"): continue
                write("  "+key+": "+val, res_dir)


        ####################################################################
        ####################################################################
        print "\n%s\n\tbuilding\n%s"%(('-'*30,)*2)
        ####################################################################
        #################################################################### 
        # ConvNet
        # ------------------------------------------------------------------------------
        # calculate resulting video shapes for all stages
        conv_shapes = []
        for i in xrange(net.n_stages):
            k,p,v = array(net.kernels[i]), array(net.pools[i]), array(tr.video_shapes[i])
            conv_s = tuple(v-k+1)
            conv_shapes.append(conv_s)
            tr.video_shapes.append(tuple((v-k+1)/p))
            print "stage", i
            if use.depth and i==0:
                print "  conv",tr.video_shapes[i],"x 2 ->",conv_s #for body and hand
            else:
                print "  conv",tr.video_shapes[i],"->",conv_s
            print "  pool",conv_s,"->",tr.video_shapes[i+1],"x",net.maps[i+1]

        # number of inputs for MLP = (# maps last stage)*(# convnets)*(resulting video shape) + trajectory size
        n_in_MLP = net.maps[-1]*net.n_convnets*prod(tr.video_shapes[-1]) 

        if use.depth:
            if net.n_convnets==2: 
                out = [x[:,:,0,:,:,:], x[:,:,1,:,:,:]] # 2 nets: body and hand

        # build 3D ConvNet
        
        for stage in xrange(net.n_stages):
            for i in xrange(len(out)): # for body and hand
                # normalization
                if use.norm and stage==0: 
                    gray_norm = NormLayer(out[i][:,0:1], method="lcn",
                        use_divisor=use.norm_div).output
                    gray_norm = std_norm(gray_norm,axis=[-3,-2,-1])
                    depth_norm = var_norm(out[i][:,1:])
                    out[i]  = T.concatenate([gray_norm,depth_norm],axis=1)
                elif use.norm:
                    out[i] = NormLayer(out[i], method="lcn",use_divisor=use.norm_div).output
                    out[i] = std_norm(out[i],axis=[-3,-2,-1])
                # convolutions  
                out[i] *= net.scaler[stage][i]
                self.layers.append(ConvLayer(out[i], **conv_args(stage, i, batch, net, use, tr.rng, tr.video_shapes, load_path)))
                out[i] = self.layers[-1].output
                out[i] = PoolLayer(out[i], net.pools[stage], method=net.pool_method).output
                if tr.inspect: 
                    self.insp_mean.append(T.mean(out[i]))
                    self.insp_std.append(T.std(out[i]))

        # flatten all convnets outputs
        for i in xrange(len(out)): out[i] = std_norm(out[i],axis=[-3,-2,-1])
        out = [out[i].flatten(2) for i in range(len(out))]
        vid_ = T.concatenate(out, axis=1)

        # dropout
        if use.drop: 
            drop.p_vid = shared(float32(drop.p_vid_val) )
            drop.p_hidden = shared(float32(drop.p_hidden_val))
            vid_ = DropoutLayer(vid_, rng=tr.rng, p=drop.p_hidden).output

        #maxout
        if use.maxout:
            vid_ = maxout(vid_, (batch.micro,n_in_MLP))
            net.activation = lin
            n_in_MLP /= 2
            # net.hidden *= 2

        # MLP
        # ------------------------------------------------------------------------------
        # fusion
        if net.fusion == "early":
            out = vid_
            # hidden layer
            if use.load:
                W,b = load_params(use, load_path)
                self.layers.append(HiddenLayer(out, n_in=n_in_MLP, n_out=net.hidden_vid, rng=tr.rng, W=W, b=b,
                    W_scale=net.W_scale[-2], b_scale=net.b_scale[-2], activation=net.activation))
            else:
                self.layers.append(HiddenLayer(out, n_in=n_in_MLP, n_out=net.hidden_vid, rng=tr.rng, 
                    W_scale=net.W_scale[-2], b_scale=net.b_scale[-2], activation=net.activation))
            out = self.layers[-1].output


        #if tr.inspect: 
            #self.insp_mean = T.stack(self.insp_mean)
            #self.insp_std = T.stack(self.insp_std)
            #self.insp = T.stack(self.insp[0],self.insp[1],self.insp[2],self.insp[3],self.insp[4],self.insp[5], T.mean(out))
        #else: self.insp =  T.stack(0,0)
        # out = normalize(out)
        if use.drop: out = DropoutLayer(out, rng=tr.rng, p=drop.p_hidden).output
        #maxout
        if use.maxout:
            out = maxout(out, (batch.micro,net.hidden))
            net.hidden /= 2

        
        # now assembly all the output
        self.out = out
        self.n_in_MLP = n_in_MLP
Example #6
0
# customized data loader for both video module and skeleton module
#loader = DataLoader_with_skeleton_normalisation(src, tr.batch_size, Mean_CNN, Std_CNN) # Lio changed it to read from HDF5 files
# we load the CNN parameteres here
x = ndtensor(len(tr.in_shape))(name='x')  # video input
x_ = _shared(empty(tr.in_shape))

use.load = True
use.fast_conv = True
video_cnn = conv3d_chalearn(x, use, lr, batch, net, reg, drop, mom, tr,
                            res_dir, load_path)

out = video_cnn.out
layers = []  # all architecture layers
# softmax layer
if use.load:
    W, b = load_params(use, load_path)
    print W.shape, b.shape
    layers.append(
        LogRegr(out,
                rng=tr.rng,
                n_in=net.hidden_vid,
                W=W,
                b=b,
                W_scale=net.W_scale[-1],
                b_scale=net.b_scale[-1],
                n_out=net.n_class))
else:
    layers.append(
        LogRegr(out,
                rng=tr.rng,
                n_in=net.hidden_vid,
# customized data loader for both video module and skeleton module
#loader = DataLoader_with_skeleton_normalisation(src, tr.batch_size, Mean_CNN, Std_CNN) # Lio changed it to read from HDF5 files
# we load the CNN parameteres here
x = ndtensor(len(tr.in_shape))(name = 'x') # video input
x_ = _shared(empty(tr.in_shape))


use.load=True
use.fast_conv=True
video_cnn = conv3d_chalearn(x, use, lr, batch, net, reg, drop, mom, tr, res_dir, load_path)

out = video_cnn.out
layers = [] # all architecture layers
# softmax layer
if use.load:
    W, b = load_params(use, load_path)
    print W.shape, b.shape
    layers.append(LogRegr(out, rng=tr.rng, n_in=net.hidden_vid, W=W, b=b,
        W_scale=net.W_scale[-1], b_scale=net.b_scale[-1], n_out=net.n_class))
else:
    layers.append(LogRegr(out, rng=tr.rng, n_in=net.hidden_vid, 
        W_scale=net.W_scale[-1], b_scale=net.b_scale[-1], n_out=net.n_class))

print "compiling output"
test_model = function([], layers[-1].p_y_given_x, 
                      givens={x: x_}, 
                    on_unused_input='ignore')

os.chdir(data)
samples=glob("*.zip") 
print len(samples), "samples found"
#maxout
if use.maxout:
    vid_ = maxout(vid_, (batch.micro,n_in_MLP))
    net.activation = lin
    n_in_MLP /= 2
    # net.hidden *= 2

# MLP
# ------------------------------------------------------------------------------
# fusion
if net.fusion == "early":
    out = vid_
    # hidden layer
    if use.load:
        Wh, bh = load_params(use)  # This is test, wudi added this!
        layers.append(HiddenLayer(out, W = Wh, b =bh, n_in=n_in_MLP, n_out=net.hidden, rng=tr.rng, 
            W_scale=net.W_scale[-2], b_scale=net.b_scale[-2], activation=leaky_relu))
    else:
        layers.append(HiddenLayer(out, n_in=n_in_MLP, n_out=net.hidden, rng=tr.rng, 
            W_scale=net.W_scale[-2], b_scale=net.b_scale[-2], activation=leaky_relu))
    out = layers[-1].output


if tr.inspect: insp = T.stack(insp[0],insp[1],insp[2],insp[3],insp[4],insp[5], T.mean(out))
else: insp =  T.stack(0,0)
# out = normalize(out)
if use.drop: out = DropoutLayer(out, rng=tr.rng, p=drop.p_hidden).output
#maxout
if use.maxout:
    out = maxout(out, (batch.micro,net.hidden))