Skip to content
Open
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
78 changes: 57 additions & 21 deletions src/datasets.py
Original file line number Diff line number Diff line change
Expand Up @@ -317,46 +317,82 @@ def rectify_events(self, x: np.ndarray, y: np.ndarray):
return rectify_map[y, x]

def get_data(self, index) -> Dict[str, any]:
ts_start: int = self.timestamps_flow[index] - self.delta_t_us
ts_end: int = self.timestamps_flow[index]
#ts_start: int = self.timestamps_flow[index] - self.delta_t_us
#ts_end: int = self.timestamps_flow[index]
ts_start1: int = self.timestamps_flow[index] - self.delta_t_us
ts_end1: int = self.timestamps_flow[index]
ts_start2: int = self.timestamps_flow[index + 1] - self.delta_t_us
ts_end2: int = self.timestamps_flow[index + 1]

file_index = self.indices[index]
#file_index = self.indices[index]
file_index1 = self.indices[index]
file_index2 = self.indices[index + 1]

output = {
'file_index': file_index,
'timestamp': self.timestamps_flow[index],
#'file_index': file_index,
'file_index1': file_index1,
'file_index2': file_index2,
#'timestamp': self.timestamps_flow[index],
'timestamp1': self.timestamps_flow[index],
'timestamp2': self.timestamps_flow[index + 1],
'seq_name': self.seq_name
}
# Save sample for benchmark submission
output['save_submission'] = file_index in self.idx_to_visualize
output['visualize'] = self.visualize_samples
event_data = self.event_slicer.get_events(
ts_start, ts_end)
p = event_data['p']
t = event_data['t']
x = event_data['x']
y = event_data['y']

#event_data = self.event_slicer.get_events(ts_start, ts_end)
event_data1 = self.event_slicer.get_events(ts_start1, ts_end1)
event_data2 = self.event_slicer.get_events(ts_start2, ts_end2)

#p = event_data['p']
#t = event_data['t']
#x = event_data['x']
#y = event_data['y']

#event_data1 = self.event_slicer.get_events(ts_start1, ts_end1)
p1 = event_data1['p']
t1 = event_data1['t']
x1 = event_data1['x']
y1 = event_data1['y']

#event_data2 = self.event_slicer.get_events(ts_start2, ts_end2)
p2 = event_data2['p']
t2 = event_data2['t']
x2 = event_data2['x']
y2 = event_data2['y']

p = np.concatenate((p1, p2))
t = np.concatenate((t1, t2))
x = np.concatenate((x1, x2))
y = np.concatenate((y1, y2))


xy_rect = self.rectify_events(x, y)
x_rect = xy_rect[:, 0]
y_rect = xy_rect[:, 1]

if self.voxel_grid is None:
raise NotImplementedError
else:
event_representation = self.events_to_voxel_grid(
p, t, x_rect, y_rect)
event_representation = self.events_to_voxel_grid(p, t, x_rect, y_rect)
output['event_volume'] = event_representation
output['name_map'] = self.name_idx

if self.load_gt:
output['flow_gt'
] = [torch.tensor(x) for x in self.load_flow(self.flow_png[index])]

output['flow_gt'
][0] = torch.moveaxis(output['flow_gt'][0], -1, 0)
output['flow_gt'
][1] = torch.unsqueeze(output['flow_gt'][1], 0)
if self.load_gt:
#output['flow_gt'] = [torch.tensor(x) for x in self.load_flow(self.flow_png[index])]
flow_gt1 = [torch.tensor(x) for x in self.load_flow(self.flow_png[index])]
flow_gt2 = [torch.tensor(x) for x in self.load_flow(self.flow_png[index + 1])]

#output['flow_gt'][0] = torch.moveaxis(output['flow_gt'][0], -1, 0)
#output['flow_gt'][1] = torch.unsqueeze(output['flow_gt'][1], 0)
flow_gt1[0] = torch.moveaxis(flow_gt1[0], -1, 0)
flow_gt1[1] = torch.unsqueeze(flow_gt1[1], 0)
flow_gt2[0] = torch.moveaxis(flow_gt2[0], -1, 0)
flow_gt2[1] = torch.unsqueeze(flow_gt2[1], 0)

output['flow_gt'] = [flow_gt1, flow_gt2]

return output

def __getitem__(self, idx):
Expand Down