Skip to content

Add object detection, tracking, mobile notifications, and search to any security camera.

clean

roryclearbf9ba347d88f
diff --git a/detection/rfdetr.py b/detection/rfdetr.py
index 0b5acfc..0262118 100644
--- a/detection/rfdetr.py
+++ b/detection/rfdetr.py
@@ -266,21 +266,19 @@ def gen_sineembed_for_position(pos_tensor, dim=128):
pos = Tensor.cat(pos_y, pos_x, pos_w, pos_h, dim=2)
return pos
-class TransformerDecoder(): # todo remove unused
+class TransformerDecoder():
def __call__(self, tgt, memory, memory_key_padding_mask=None,
refpoints_unsigmoid=None, level_start_index=None, spatial_shapes=None):
intermediate = []
def get_reference(refpoints_unsigmoid):
obj_center = refpoints_unsigmoid[..., :4]
refpoints_input = obj_center[:, :, None]
- query_sine_embed = gen_sineembed_for_position(
- refpoints_input[:, :, 0, :], 256 / 2) # bs, nq, 256*2
+ query_sine_embed = gen_sineembed_for_position(refpoints_input[:, :, 0, :], 256 / 2)
query_pos = self.ref_point_head(query_sine_embed)
return refpoints_input, query_pos
- for _, layer in enumerate(self.layers):
- refpoints_input, query_pos = get_reference(refpoints_unsigmoid) #todo
-
+ refpoints_input, query_pos = get_reference(refpoints_unsigmoid)
+ for layer in self.layers:
tgt = layer(tgt, memory,
memory_key_padding_mask=memory_key_padding_mask,
query_pos=query_pos,
@@ -291,7 +289,7 @@ class TransformerDecoder(): # todo remove unused
x = self.norm(tgt)
intermediate.append(x)
return [(Tensor.stack(intermediate)), (refpoints_unsigmoid.unsqueeze(0))]
-
+
def gen_encoder_output_proposals(memory, memory_padding_mask, spatial_shape, unsigmoid=True):
memory_padding_mask = memory_padding_mask.cast(dtype=dtypes.bool)
@@ -399,7 +397,6 @@ class Bottleneck():
class C2f():
def __init__(self): self.c = 128
def __call__(self, x):
- """Forward pass using split() instead of chunk()."""
y = list(self.cv1(x).split((self.c, self.c), 1))
y.extend(m(y[-1]) for m in self.m)
y = Tensor.cat(*y, dim=1)