wire up FFN model — model_def, label map fix, TF requirement

.gitignore

@@ -0,0 +1,4 @@
+model.keras
+*.keras
+__pycache__/
+.DS_Store

requirements.txt

@@ -2,3 +2,4 @@ streamlit
 pandas
 numpy
 scipy
+tensorflow-cpu==2.16.2

src/model_def.py

@@ -0,0 +1,65 @@
+"""FeedForwardNetwork definition required for deserializing model.keras.
+
+This must be imported before tf.keras.models.load_model() is called so
+that keras can resolve the registered custom class.
+"""
+
+import keras
+import tensorflow as tf
+
+
+@keras.saving.register_keras_serializable()
+class FeedForwardNetwork(tf.keras.Model):
+    """Fully-connected feedforward network for 6-class HAR classification.
+
+    Architecture: Dense(512) → BN → ReLU → Dropout
+                  Dense(256) → BN → ReLU → Dropout
+                  Dense(128) → BN → ReLU → Dropout
+                  Dense(6, softmax)
+    """
+
+    def __init__(
+        self,
+        num_features,
+        num_classes,
+        hidden_units=(512, 256, 128),
+        dropout_rate=0.3,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self._num_features = num_features
+        self._num_classes = num_classes
+        self._hidden_units = tuple(hidden_units)
+        self._dropout_rate = dropout_rate
+
+        self.hidden_blocks = []
+        for units in hidden_units:
+            self.hidden_blocks.append([
+                tf.keras.layers.Dense(units, use_bias=False),
+                tf.keras.layers.BatchNormalization(),
+                tf.keras.layers.ReLU(),
+                tf.keras.layers.Dropout(dropout_rate),
+            ])
+
+        self.output_layer = tf.keras.layers.Dense(num_classes, activation="softmax")
+
+    def call(self, inputs, training=False):
+        x = inputs
+        for block in self.hidden_blocks:
+            for layer in block:
+                if isinstance(layer, (tf.keras.layers.BatchNormalization,
+                                      tf.keras.layers.Dropout)):
+                    x = layer(x, training=training)
+                else:
+                    x = layer(x)
+        return self.output_layer(x)
+
+    def get_config(self):
+        config = super().get_config()
+        config.update({
+            "num_features": self._num_features,
+            "num_classes": self._num_classes,
+            "hidden_units": self._hidden_units,
+            "dropout_rate": self._dropout_rate,
+        })
+        return config

src/streamlit_app.py

@@ -5,12 +5,12 @@ import pandas as pd
 # ── Constants ──────────────────────────────────────────────────────────────
 
 LABEL_MAP = {
-    0: "LAYING",
-    1: "SITTING",
-    2: "STANDING",
-    3: "WALKING",
-    4: "WALKING_DOWNSTAIRS",
-    5: "WALKING_UPSTAIRS",
+    0: "WALKING",
+    1: "WALKING_UPSTAIRS",
+    2: "WALKING_DOWNSTAIRS",
+    3: "SITTING",
+    4: "STANDING",
+    5: "LAYING",
 }
 
 ACTIVITY_ICONS = {
@@ -35,7 +35,16 @@ EXPLANATIONS = {
 
 @st.cache_resource
 def load_model():
-    return None, "no_model"
+    try:
+        import tensorflow as tf
+        from model_def import FeedForwardNetwork
+        model = tf.keras.models.load_model(
+            "model.keras",
+            custom_objects={"FeedForwardNetwork": FeedForwardNetwork},
+        )
+        return model, "ready"
+    except Exception:
+        return None, "no_model"
 
 # ── Page config ─────────────────────────────────────────────────────────────
 
@@ -66,8 +75,8 @@ with st.sidebar:
     """)
     st.markdown("---")
     st.markdown("**Model performance on test set**")
-    st.metric("Accuracy", "—")
-    st.metric("Macro F1", "—")
+    st.metric("Architecture", "FFN 512→256→128")
+    st.metric("Status", "FFN live · CNN pending")
     st.markdown("---")
     st.caption("DAT606 Group Assignment · Pan-Atlantic University")