Add cuda option and read binary mode for pickle tracks.pckl

SyncNetInstance.py

@@ -12,6 +12,7 @@
 from scipy.io import wavfile
 from SyncNetModel import *
 
+cuda = False
 
 # ==================== Get OFFSET ====================
 
@@ -101,12 +102,18 @@ def evaluate(self, opt, videofile):
 
             im_batch = [ imtv[:,:,vframe:vframe+5,:,:] for vframe in range(i,min(lastframe,i+opt.batch_size)) ]
             im_in = torch.cat(im_batch,0)
-            im_out  = self.__S__.forward_lip(im_in.cuda());
+            if cuda:
+                im_out  = self.__S__.forward_lip(im_in.cuda());
+            else:
+                im_out  = self.__S__.forward_lip(im_in);
             im_feat.append(im_out.data.cpu())
 
             cc_batch = [ cct[:,:,:,vframe*4:vframe*4+20] for vframe in range(i,min(lastframe,i+opt.batch_size)) ]
             cc_in = torch.cat(cc_batch,0)
-            cc_out  = self.__S__.forward_aud(cc_in.cuda())
+            if cuda:
+                cc_out  = self.__S__.forward_aud(cc_in.cuda())
+            else:
+                cc_out  = self.__S__.forward_aud(cc_in)
             cc_feat.append(cc_out.data.cpu())
 
         im_feat = torch.cat(im_feat,0)

SyncNetModel.py

@@ -1,6 +1,8 @@
 import torch
 import torch.nn as nn
 
+cuda = False
+
 def save(model, filename):
     with open(filename, "wb") as f:
         torch.save(model, f);
@@ -90,11 +92,11 @@ def __init__(self, num_layers_in_fc_layers = 1024):
             nn.ReLU(inplace=True),
         );
 
-        self.netcnnaud = self.netcnnaud.cuda();
-        self.netcnnlip = self.netcnnlip.cuda();
-        self.netfcaud  = self.netfcaud.cuda();
-        self.netfclip = self.netfclip.cuda();
-
+        if cuda:
+            self.netcnnaud = self.netcnnaud.cuda();
+            self.netcnnlip = self.netcnnlip.cuda();
+            self.netfcaud  = self.netfcaud.cuda();
+            self.netfclip = self.netfclip.cuda();
 
     def forward_aud(self, x):
 
@@ -110,4 +112,4 @@ def forward_lip(self, x):
         mid = mid.view((mid.size()[0], -1)); # N x (ch x 24)
         out = self.netfclip(mid);
 
-        return out;
+        return out;

run_syncnet.py

@@ -31,7 +31,7 @@
 
 # ==================== GET OFFSETS ====================
 
-with open(os.path.join(opt.work_dir,opt.reference,'tracks.pckl'), 'r') as fil:
+with open(os.path.join(opt.work_dir,opt.reference,'tracks.pckl'), 'rb') as fil:
     tracks = pickle.load(fil)
 
 dists = []