MCPcopy Index your code
hub / github.com/FoundationVision/ByteTrack / eval_mota

Function eval_mota

tools/interpolation.py:14–30  ·  view source on GitHub ↗
(data_root, txt_path)

Source from the content-addressed store, hash-verified

12
13
14def eval_mota(data_root, txt_path):
15 accs = []
16 seqs = sorted([s for s in os.listdir(data_root) if s.endswith('FRCNN')])
17 #seqs = sorted([s for s in os.listdir(data_root)])
18 for seq in seqs:
19 video_out_path = os.path.join(txt_path, seq + '.txt')
20 evaluator = Evaluator(data_root, seq, 'mot')
21 accs.append(evaluator.eval_file(video_out_path))
22 metrics = mm.metrics.motchallenge_metrics
23 mh = mm.metrics.create()
24 summary = Evaluator.get_summary(accs, seqs, metrics)
25 strsummary = mm.io.render_summary(
26 summary,
27 formatters=mh.formatters,
28 namemap=mm.io.motchallenge_metric_names
29 )
30 print(strsummary)
31
32
33def get_mota(data_root, txt_path):

Callers 1

interpolation.pyFile · 0.85

Calls 3

eval_fileMethod · 0.95
EvaluatorClass · 0.90
get_summaryMethod · 0.45

Tested by

no test coverage detected