diff --git a/.gitignore b/.gitignore
index ee64372..188d871 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,2 +1,4 @@
 .idea/
 *.pyc
+.ipynb_checkpoints/
+__pycache__/
diff --git a/.ipynb_checkpoints/demo-checkpoint.ipynb b/.ipynb_checkpoints/demo-checkpoint.ipynb
deleted file mode 100644
index a82cd58..0000000
--- a/.ipynb_checkpoints/demo-checkpoint.ipynb
+++ /dev/null
@@ -1,163 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 26,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import matplotlib.pyplot as plt\n",
-    "import mdp.offroad_grid as offroad_grid\n",
-    "import numpy as np\n",
-    "from network.hybrid_dilated import HybridDilated\n",
-    "from torch.autograd import Variable\n",
-    "import torch\n",
-    "from os.path import join\n",
-    "import scipy.io as sio\n",
-    "from loader.util import leastsq_circle, calc_sign"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 27,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "/home/yf/git_repo/vehicle-motion-forecasting/network/hybrid_dilated.py:59: UserWarning: nn.init.kaiming_normal is now deprecated in favor of nn.init.kaiming_normal_.\n",
-      "  nn.init.kaiming_normal(mod.weight, a=0)\n"
-     ]
-    }
-   ],
-   "source": [
-    "# initialize parameters\n",
-    "grid_size = 80\n",
-    "discount = 0.9\n",
-    "model = offroad_grid.OffroadGrid(grid_size, discount)\n",
-    "n_states = model.n_states\n",
-    "n_actions = model.n_actions\n",
-    "\n",
-    "net = HybridDilated(feat_out_size=25, regression_hidden_size=64)\n",
-    "net.init_weights()\n",
-    "net.load_state_dict(torch.load(join('example_data', 'example_weights.pth'))['net_state'])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 28,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def load(grid_size):\n",
-    "    \"\"\" load sample demo input data\"\"\"\n",
-    "    mean_std = sio.loadmat(join('example_data', 'data_mean_std.mat'))\n",
-    "    data_mat = sio.loadmat(join('example_data', 'demo_input.mat'))\n",
-    "    feat = data_mat['feat']\n",
-    "    # pre-process environment input\n",
-    "    feat[0] = (feat[0] - np.mean(feat[0])) / np.std(feat[0]) # normalize max-height feature locally (w.r.t. robot frame)\n",
-    "    feat[1] = (feat[1] - mean_std['variance_mean']) / mean_std['variance_std']\n",
-    "    feat[2] = (feat[2] - mean_std['red_mean']) / mean_std['red_std']\n",
-    "    feat[3] = (feat[3] - mean_std['green_mean']) / mean_std['green_std']\n",
-    "    feat[4] = (feat[4] - mean_std['blue_mean']) / mean_std['blue_std']\n",
-    "    \n",
-    "    # pre-process kinematic input\n",
-    "    past_traj, future_traj = data_mat['past_traj'], data_mat['future_traj']\n",
-    "    x, y = past_traj[:, 0], past_traj[:, 1]\n",
-    "    xc, yc, r, _ = leastsq_circle(x, y)\n",
-    "    curve_sign = calc_sign(x[0], y[0], x[-1], y[-1], xc, yc)\n",
-    "    kappa = 1.0 / r * curve_sign * 10.0 # 10.0 is empirically selected by observing the histogram\n",
-    "    feat = np.vstack((feat, np.full((1, grid_size, grid_size), kappa, dtype=np.float)))\n",
-    "    \n",
-    "    normalization = 0.5 * grid_size # 0.5*grid_size is used for normalize vx, vy, coordinate layers\n",
-    "    vx = (past_traj[-1, 0] - past_traj[0, 0]) / normalization\n",
-    "    vy = (past_traj[-1, 1] - past_traj[0 ,1]) / normalization\n",
-    "    feat = np.vstack((feat, np.full((1, grid_size, grid_size), vx, dtype=np.float)))\n",
-    "    feat = np.vstack((feat, np.full((1, grid_size, grid_size), vy, dtype=np.float)))\n",
-    "\n",
-    "    # coordinate layer\n",
-    "    center_idx = grid_size / 2\n",
-    "    delta_x_layer = np.zeros((1, grid_size, grid_size), dtype=np.float)\n",
-    "    delta_y_layer = delta_x_layer.copy()\n",
-    "    for x in range(grid_size):\n",
-    "        for y in range(grid_size):\n",
-    "            delta_x_layer[0, x, y] = x - center_idx\n",
-    "            delta_y_layer[0, x, y] = y - center_idx\n",
-    "    feat = np.vstack((feat, delta_x_layer / normalization))\n",
-    "    feat = np.vstack((feat, delta_y_layer / normalization))\n",
-    "    \n",
-    "    return feat, past_traj, future_traj"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 31,
-   "metadata": {},
-   "outputs": [
-    {
-     "ename": "RuntimeError",
-     "evalue": "Expected object of type torch.DoubleTensor but found type torch.FloatTensor for argument #2 'weight'",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
-      "\u001b[0;32m<ipython-input-31-bc765b89d6b9>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mfeat\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexpand_dims\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfeat\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mfeat_var\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mVariable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfrom_numpy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfeat\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mDoubleTensor\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0mr_var\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnet\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfeat_var\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
-      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    489\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    490\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 491\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    492\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mhook\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    493\u001b[0m             \u001b[0mhook_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m~/git_repo/vehicle-motion-forecasting/network/hybrid_dilated.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, x)\u001b[0m\n\u001b[1;32m     44\u001b[0m         \u001b[0;31m# geometric and semantic feature extraction\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     45\u001b[0m         \u001b[0mfeat_in\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfeat_in_size\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 46\u001b[0;31m         \u001b[0mfeat_out\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfeat_block\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfeat_in\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     47\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     48\u001b[0m         \u001b[0mkinematic_in\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfeat_out\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m5\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdim\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    489\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    490\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 491\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    492\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mhook\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    493\u001b[0m             \u001b[0mhook_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/torch/nn/modules/container.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m     89\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     90\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mmodule\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_modules\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 91\u001b[0;31m             \u001b[0minput\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     92\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     93\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/torch/nn/modules/module.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *input, **kwargs)\u001b[0m\n\u001b[1;32m    489\u001b[0m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_slow_forward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    490\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 491\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    492\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mhook\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_forward_hooks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    493\u001b[0m             \u001b[0mhook_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;32m~/anaconda3/lib/python3.6/site-packages/torch/nn/modules/conv.py\u001b[0m in \u001b[0;36mforward\u001b[0;34m(self, input)\u001b[0m\n\u001b[1;32m    299\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mforward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0minput\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    300\u001b[0m         return F.conv2d(input, self.weight, self.bias, self.stride,\n\u001b[0;32m--> 301\u001b[0;31m                         self.padding, self.dilation, self.groups)\n\u001b[0m\u001b[1;32m    302\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    303\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
-      "\u001b[0;31mRuntimeError\u001b[0m: Expected object of type torch.DoubleTensor but found type torch.FloatTensor for argument #2 'weight'"
-     ]
-    }
-   ],
-   "source": [
-    "feat, past_traj, future_traj = load(grid_size)\n",
-    "feat = np.expand_dims(feat, axis=0)\n",
-    "feat_var = Variable(torch.from_numpy(feat).type(torch.DoubleTensor))\n",
-    "r_var = net(feat_var)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 30,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "torch.from_numpy??"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "collapsed": true
-   },
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.6.3"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 2
-}
diff --git a/demo.ipynb b/demo.ipynb
index 4fa8980..7c81457 100644
--- a/demo.ipynb
+++ b/demo.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -15,12 +15,13 @@
     "from os.path import join\n",
     "import scipy.io as sio\n",
     "from loader.util import leastsq_circle, calc_sign\n",
-    "import seaborn as sns"
+    "import seaborn as sns\n",
+    "import viz"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -61,14 +62,14 @@
        ")"
       ]
      },
-     "execution_count": 9,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
     "# initialize parameters\n",
-    "grid_size = 80\n",
+    "grid_size = 80 # local grid environment size\n",
     "discount = 0.9\n",
     "model = offroad_grid.OffroadGrid(grid_size, discount)\n",
     "n_states = model.n_states\n",
@@ -82,14 +83,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
-    "def load(grid_size):\n",
+    "def load(grid_size, fname):\n",
     "    \"\"\" load sample demo input data\"\"\"\n",
     "    mean_std = sio.loadmat(join('example_data', 'data_mean_std.mat'))\n",
-    "    data_mat = sio.loadmat(join('example_data', 'demo_input.mat'))\n",
+    "    data_mat = sio.loadmat(join('example_data', fname))\n",
     "    feat = data_mat['feat']\n",
     "    # pre-process environment input\n",
     "    feat[0] = (feat[0] - np.mean(feat[0])) / np.std(feat[0]) # normalize max-height feature locally (w.r.t. robot frame)\n",
@@ -128,28 +129,27 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 14,
    "metadata": {
     "scrolled": false
    },
    "outputs": [],
    "source": [
-    "feat, past_traj, future_traj = load(grid_size)\n",
-    "feat = np.expand_dims(feat, axis=0)\n",
-    "feat_var = Variable(torch.from_numpy(feat).float())\n",
-    "r_var = net(feat_var)"
+    "feat, past_traj, future_traj = load(grid_size, 'example_input_narrow_trail.mat')\n",
+    "feat_var = Variable(torch.from_numpy(np.expand_dims(feat, axis=0)).float())\n",
+    "r_var = net(feat_var) # forward reward inference"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
      "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAD/CAYAAADPJgxuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkQqiqIAjRYoUqSD6FyOFgiJ6v7gyAAAAAElFTkSuQmCC\n",
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\neEhPBgAAAABJRU5ErkJggg==\n",
       "text/plain": [
-       "<matplotlib.figure.Figure at 0x7f7bf80c64a8>"
+       "<matplotlib.figure.Figure at 0x7f00ba085630>"
       ]
      },
      "metadata": {},
@@ -157,43 +157,23 @@
     }
    ],
    "source": [
-    "r = r_var[0].data.numpy().squeeze()\n",
-    "sns.heatmap(r, cmap='viridis')\n",
+    "# visualize input: traj overlayed with top-down RGB\n",
+    "rgb = viz.feat2rgb(feat)\n",
+    "rgb_with_traj = viz.overlay(rgb, future_traj, past_traj)\n",
+    "plt.imshow(rgb_with_traj)\n",
     "plt.show()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 24,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "find_optimal_value. iteration 17, last update 0.09115273836986937\n"
-     ]
-    }
-   ],
-   "source": [
-    "r_vector = r.reshape(n_states) # convert 2D reward matrix to a 1D vector\n",
-    "value_vector = model.find_optimal_value(r_vector, 0.1)\n",
-    "policy = model.find_stochastic_policy(value_vector, r_vector)\n",
-    "past_traj_len = past_traj.shape[0]\n",
-    "svf_vector = model.find_svf_demo(policy, past_traj_len)\n",
-    "svf = np.log(svf_vector.reshape(grid_size, grid_size) + 1e-3)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 25,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
      "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWgAAAD/CAYAAADc8UyaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQTuO4+QU76Adx3Fyyv8HLv8I9GZvl3kAAAAASUVORK5CYII=\n",
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAD/CAYAAADPJgxuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text/plain": [
-       "<matplotlib.figure.Figure at 0x7f7bfadb30f0>"
+       "<matplotlib.figure.Figure at 0x7f00b9fed438>"
       ]
      },
      "metadata": {},
@@ -201,63 +181,28 @@
     }
    ],
    "source": [
-    "sns.heatmap(svf, cmap='viridis')\n",
+    "r = r_var[0].data.numpy().squeeze()\n",
+    "sns.heatmap(r, cmap='viridis')\n",
     "plt.show()"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 30,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "xlayer = np.zeros((80,80))\n",
-    "ylayer = xlayer.copy()\n",
-    "for x in range(80):\n",
-    "    for y in range(80):\n",
-    "        xlayer[x,y] = x - 40\n",
-    "        ylayer[x,y] = y - 40"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 47,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
-     "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoMAAAJyCAYAAAC/oMoyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQQCABtMMAgA0mGYQAKDBNIMAAA32/wCVml2x6abQHwAAAABJRU5ErkJggg==\n",
-      "text/plain": [
-       "<matplotlib.figure.Figure at 0x7f7bfa96eac8>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "fig=plt.figure(figsize=(10, 10), dpi= 80, facecolor='w', edgecolor='k')\n",
-    "plt.gca().set_aspect('equal', adjustable='box')\n",
-    "# plt.xticks(np.arange(0,80,10), fontsize=20)\n",
-    "# plt.yticks(np.arange(0,80,10), fontsize=20)\n",
-    "ax = sns.heatmap(xlayer/80, cmap='viridis')\n",
-    "ax.set_xticklabels([])\n",
-    "ax.set_yticklabels([])\n",
-    "ax.set_xlabel('Y', fontsize=20)\n",
-    "ax.set_ylabel('X', fontsize=20)\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 46,
-   "metadata": {},
-   "outputs": [
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "find_optimal_value. iteration 38, last update 0.00961954428426015\n"
+     ]
+    },
     {
      "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAn8AAAJyCAYAAAC10KYlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+X/u9QC+4zVazAAAAABJRU5ErkJggg==\n",
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWgAAAD/CAYAAADc8UyaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMgAAAABJRU5ErkJggg==\n",
       "text/plain": [
-       "<matplotlib.figure.Figure at 0x7f7bf80b4748>"
+       "<matplotlib.figure.Figure at 0x7f00cd2d6b00>"
       ]
      },
      "metadata": {},
@@ -265,15 +210,13 @@
     }
    ],
    "source": [
-    "fig=plt.figure(figsize=(10, 10), dpi= 80, facecolor='w', edgecolor='k')\n",
-    "plt.gca().set_aspect('equal', adjustable='box')\n",
-    "# plt.xticks(np.arange(0,80,10), fontsize=20)\n",
-    "# plt.yticks(np.arange(0,80,10), fontsize=20)\n",
-    "ax = sns.heatmap(ylayer, cmap='viridis')\n",
-    "ax.set_xticklabels([])\n",
-    "ax.set_yticklabels([])\n",
-    "ax.set_xlabel('Y', fontsize=20)\n",
-    "ax.set_ylabel('X', fontsize=20)\n",
+    "r_vector = r.reshape(n_states) # convert 2D reward matrix to a 1D vector\n",
+    "value_vector = model.find_optimal_value(r_vector, 0.01)\n",
+    "policy = model.find_stochastic_policy(value_vector, r_vector)\n",
+    "past_traj_len = past_traj.shape[0]\n",
+    "svf_vector = model.find_svf_demo(policy, past_traj_len)\n",
+    "svf = np.log(svf_vector.reshape(grid_size, grid_size) + 1e-3)\n",
+    "sns.heatmap(svf, cmap='viridis')\n",
     "plt.show()"
    ]
   },
diff --git a/demo.py b/demo.py
index 93edcc0..1ac7ef9 100644
--- a/demo.py
+++ b/demo.py
@@ -1,110 +1,86 @@
-import matplotlib
-matplotlib.use('Agg')
 import matplotlib.pyplot as plt
-
 import mdp.offroad_grid as offroad_grid
-import loader.offroad_loader as offroad_loader
-from torch.utils.data import DataLoader
 import numpy as np
-
-from network.hybrid_fcn import HybridFCN
 from network.hybrid_dilated import HybridDilated
-
 from torch.autograd import Variable
 import torch
-import os
-from tqdm import tqdm
+from os.path import join
 import scipy.io as sio
-import imageio
-import time
-
-def overlay(img, future_traj, past_traj):
-    overlay_img = img.copy()
-    for p in future_traj:
-        overlay_img[int(p[0]), int(p[1]), 0] = 255  # red
-        overlay_img[int(p[0]), int(p[1]), 1] = 255  # green
-        overlay_img[int(p[0]), int(p[1]), 2] = 255  # blue
-    for p in past_traj:
-        overlay_img[int(p[0]), int(p[1]), 0] = 255
-        overlay_img[int(p[0]), int(p[1]), 1] = 0
-        overlay_img[int(p[0]), int(p[1]), 2] = 0
-    return overlay_img
-
+from loader.util import leastsq_circle, calc_sign
+import seaborn as sns
+import viz
 
-def pred(feat, future_traj, net, n_states, model, grid_size, past_traj):
-    # n_sample = feat.shape[0]
-    # start = time.clock()
-    feat = feat.float()
-    feat_var = Variable(feat)
-    r_var = net(feat_var)
 
-    r_sample = r_var[0].data.numpy().squeeze().reshape(n_states)
-    future_traj_sample = future_traj[0].numpy()  # choose one sample from the batch
-    future_traj_sample = future_traj_sample[~np.isnan(future_traj_sample).any(axis=1)]  # remove appended NAN rows
-    future_traj_sample = future_traj_sample.astype(np.int64)
-    past_traj_sample = past_traj[0].numpy()  # choose one sample from the batch
-    past_traj_sample = past_traj_sample[~np.isnan(past_traj_sample).any(axis=1)]  # remove appended NAN rows
-    past_traj_sample = past_traj_sample.astype(np.int64)
-
-    values_sample = model.find_optimal_value(r_sample, 0.1)
-    policy = model.find_stochastic_policy(values_sample, r_sample)
-    svf_sample = model.find_svf_demo(future_traj_sample, policy, past_traj_sample)
-    # print('{} s'.format(time.clock()-start))
-    svf = svf_sample.reshape(grid_size, grid_size)
-    reward = r_var.data[0, 0].numpy()
-    return reward, svf
-
-
-# initialize param
+# initialize parameters
 grid_size = 80
 discount = 0.9
-
-exp = '6.34'
-resume='step940-loss0.720122159457321.pth'
-net = HybridDilated(feat_out_size=25, regression_hidden_size=64)
-
 model = offroad_grid.OffroadGrid(grid_size, discount)
 n_states = model.n_states
 n_actions = model.n_actions
 
-num = '2.1'
-
-loader = offroad_loader.OffroadLoader(grid_size=grid_size, train=False, demo='demo_data_{}'.format(num), tangent=False)
-#loader = offroad_loader.OffroadLoader(grid_size=grid_size, train=False, tangent=False)
-loader.data_list.sort()
-data_list = loader.data_list
-loader = DataLoader(loader, num_workers=1, batch_size=1, shuffle=False)
-data_normalization = sio.loadmat('/data/datasets/yanfu/irl_data/train-data-mean-std.mat')
-
+net = HybridDilated(feat_out_size=25, regression_hidden_size=64)
 net.init_weights()
-checkpoint = torch.load(os.path.join('exp', exp, resume))
-net.load_state_dict(checkpoint['net_state'])
+net.load_state_dict(torch.load(join('example_data', 'example_weights6.34.pth'))['net_state'])
 net.eval()
 
 
-root = os.path.join('paper_demo_viz_{}'.format(num), exp)
-print(root)
-
-if not os.path.exists(root):
-    os.makedirs(root)
-
-for step, (feat, past_traj, future_traj) in enumerate(loader):
-    start = time.clock()
-    feat = feat.float()
-    feat_var = Variable(feat)
-    r_var = net(feat_var)
-    r_sample = r_var[0].data.numpy().squeeze().reshape(n_states)
-    print('{} s: reward network'.format(time.clock()-start))
-
-    start2 = time.clock()
-    values_sample = model.find_optimal_value(r_sample, 0.5)
-    policy = model.find_stochastic_policy(values_sample, r_sample)
-    future_traj_sample = future_traj[0].numpy()  # choose one sample from the batch
-    future_traj_sample = future_traj_sample[~np.isnan(future_traj_sample).any(axis=1)]  # remove appended NAN rows
-    future_traj_sample = future_traj_sample.astype(np.int64)
-    past_traj_sample = past_traj[0].numpy()  # choose one sample from the batch
-    past_traj_sample = past_traj_sample[~np.isnan(past_traj_sample).any(axis=1)]  # remove appended NAN rows
-    past_traj_sample = past_traj_sample.astype(np.int64)
-    svf_sample = model.find_svf_demo(policy, past_traj_sample.shape[0])
-    svf = svf_sample.reshape(grid_size, grid_size)
-    print('{} s: find svf'.format(time.clock()-start2))
+def load(grid_size):
+    """ load sample demo input data"""
+    mean_std = sio.loadmat(join('example_data', 'data_mean_std.mat'))
+    data_mat = sio.loadmat(join('example_data', 'demo_input.mat'))
+    feat = data_mat['feat']
+    # pre-process environment input
+    feat[0] = (feat[0] - np.mean(feat[0])) / np.std(feat[0])  # normalize max-height feature locally (w.r.t. robot frame)
+    feat[1] = (feat[1] - mean_std['variance_mean']) / mean_std['variance_std']
+    feat[2] = (feat[2] - mean_std['red_mean']) / mean_std['red_std']
+    feat[3] = (feat[3] - mean_std['green_mean']) / mean_std['green_std']
+    feat[4] = (feat[4] - mean_std['blue_mean']) / mean_std['blue_std']
+
+    # pre-process kinematic input
+    past_traj, future_traj = data_mat['past_traj'], data_mat['future_traj']
+    x, y = past_traj[:, 0], past_traj[:, 1]
+    xc, yc, r, _ = leastsq_circle(x, y)
+    curve_sign = calc_sign(x[0], y[0], x[-1], y[-1], xc, yc)
+    kappa = 1.0 / r * curve_sign * 10.0  # 10.0 is empirically selected by observing the histogram
+    feat = np.vstack((feat, np.full((1, grid_size, grid_size), kappa, dtype=np.float)))
+
+    normalization = 0.5 * grid_size  # 0.5*grid_size is used for normalize vx, vy, coordinate layers
+    vx = (past_traj[-1, 0] - past_traj[0, 0]) / normalization
+    vy = (past_traj[-1, 1] - past_traj[0, 1]) / normalization
+    feat = np.vstack((feat, np.full((1, grid_size, grid_size), vx, dtype=np.float)))
+    feat = np.vstack((feat, np.full((1, grid_size, grid_size), vy, dtype=np.float)))
+
+    # coordinate layer
+    center_idx = grid_size / 2
+    delta_x_layer = np.zeros((1, grid_size, grid_size), dtype=np.float)
+    delta_y_layer = delta_x_layer.copy()
+    for x in range(grid_size):
+        for y in range(grid_size):
+            delta_x_layer[0, x, y] = x - center_idx
+            delta_y_layer[0, x, y] = y - center_idx
+    feat = np.vstack((feat, delta_x_layer / normalization))
+    feat = np.vstack((feat, delta_y_layer / normalization))
+
+    return feat, past_traj, future_traj
+
+feat, past_traj, future_traj = load(grid_size)
+print(feat.shape)
+feat_var = Variable(torch.from_numpy(np.expand_dims(feat, axis=0)).float())
+r_var = net(feat_var)
+
+# visualize input and output
+rgb = viz.feat2rgb(feat)
+rgb_with_path = viz.overlay(rgb, future_traj, past_traj)
+plt.imshow(rgb_with_path)
+plt.show()
+r = r_var[0].data.numpy().squeeze()
+sns.heatmap(r, cmap='viridis')
+plt.show()
+r_vector = r.reshape(n_states) # convert 2D reward matrix to a 1D vector
+value_vector = model.find_optimal_value(r_vector, 0.005)
+policy = model.find_stochastic_policy(value_vector, r_vector)
+past_traj_len = past_traj.shape[0]
+svf_vector = model.find_svf_demo(policy, past_traj_len)
+svf = np.log(svf_vector.reshape(grid_size, grid_size) + 1e-3)
+sns.heatmap(svf, cmap='viridis')
+plt.show()
\ No newline at end of file
diff --git a/example_data/example_input_narrow_trail.mat b/example_data/example_input_narrow_trail.mat
new file mode 100644
index 0000000..7f2fe6a
Binary files /dev/null and b/example_data/example_input_narrow_trail.mat differ
diff --git a/example_data/example_weights6.33.pth b/example_data/example_weights6.33.pth
new file mode 100644
index 0000000..65064cc
Binary files /dev/null and b/example_data/example_weights6.33.pth differ
diff --git a/example_data/example_weights6.34.pth b/example_data/example_weights6.34.pth
new file mode 100644
index 0000000..f08938f
Binary files /dev/null and b/example_data/example_weights6.34.pth differ
diff --git a/mdp/__pycache__/offroad_grid.cpython-36.pyc b/mdp/__pycache__/offroad_grid.cpython-36.pyc
index a07681f..bd0c191 100644
Binary files a/mdp/__pycache__/offroad_grid.cpython-36.pyc and b/mdp/__pycache__/offroad_grid.cpython-36.pyc differ
diff --git a/mdp/offroad_grid.py b/mdp/offroad_grid.py
index db7c4aa..ab5ec0e 100644
--- a/mdp/offroad_grid.py
+++ b/mdp/offroad_grid.py
@@ -220,7 +220,7 @@ def find_stochastic_policy(self, value, reward):
                 Q[s, a] = reward[next_s] + self.discount * value[next_s]
 
         Q -= Q.max(axis=1).reshape((self.n_states, 1))  # For numerical stability
-        Q = np.exp(Q*1) / np.exp(Q*1).sum(axis=1).reshape((self.n_states, 1))  # softmax over actions
+        Q = np.exp(Q*20) / np.exp(Q*20).sum(axis=1).reshape((self.n_states, 1))  # softmax over actions
         return Q
 
     def find_optimal_value(self, reward, thresh=0.005):
diff --git a/viz.py b/viz.py
new file mode 100644
index 0000000..15af90a
--- /dev/null
+++ b/viz.py
@@ -0,0 +1,22 @@
+import numpy as np
+import scipy.io as sio
+
+def overlay(img, future_traj, past_traj):
+    overlay_img = img.copy()
+    for p in future_traj:
+        overlay_img[int(p[0]), int(p[1]), 0] = 255  # red
+        overlay_img[int(p[0]), int(p[1]), 1] = 255  # green
+        overlay_img[int(p[0]), int(p[1]), 2] = 255  # blue
+    for p in past_traj:
+        overlay_img[int(p[0]), int(p[1]), 0] = 255
+        overlay_img[int(p[0]), int(p[1]), 1] = 0
+        overlay_img[int(p[0]), int(p[1]), 2] = 0
+    return overlay_img
+
+def feat2rgb(feat):
+    normalization = sio.loadmat('example_data/data_mean_std.mat')
+    red = (feat[2] * normalization['red_std'] + normalization['red_mean']).astype(np.uint8)
+    green = (feat[3] * normalization['green_std'] + normalization['green_mean']).astype(np.uint8)
+    blue = (feat[4] * normalization['blue_std'] + normalization['blue_mean']).astype(np.uint8)
+    color = np.stack([red, green, blue], axis=2)
+    return color
\ No newline at end of file