Datasets:
Download code/dynamic_dataset/common.py from Heiheihaha17/MESH-Benchmark: direct link, hf CLI and curl.
- Browser
- Download file 3.19 kB
-
https://huggingface.co/datasets/Heiheihaha17/MESH-Benchmark/resolve/main/code/dynamic_dataset/common.py
- Command line
-
hf download hf://datasets/Heiheihaha17/MESH-Benchmark/code/dynamic_dataset/common.py
-
curl -L -o common.py https://huggingface.co/datasets/Heiheihaha17/MESH-Benchmark/resolve/main/code/dynamic_dataset/common.py
3.19 kB
| """Shared frozen protocol helpers for the full dynamic test set.""" | |
| from __future__ import annotations | |
| import gzip,hashlib,json,math,os,socket | |
| from pathlib import Path | |
| from datetime import datetime,timezone | |
| import numpy as np | |
| from dynamic_demo.common import SR,active,angular,sha,speech_activity,tokens | |
| STEP=3200 | |
| VERSION='dish_dynamic_test_v1' | |
| BRANCHES=['easy_human','easy_robot','hard_joint_a','hard_joint_b'] | |
| def now():return datetime.now(timezone.utc).isoformat() | |
| def require_cluster(): | |
| if 'debug' in socket.gethostname().lower() or not os.environ.get('KUBERNETES_SERVICE_HOST'): | |
| raise RuntimeError('Generation, rendering and full acceptance require an allocated CPU job') | |
| def uid(*parts):return hashlib.sha256(json.dumps((VERSION,)+parts,sort_keys=True,separators=(',',':')).encode()).hexdigest()[:24] | |
| def read(path):return json.loads(Path(path).read_text()) | |
| def rows(path): | |
| opener=gzip.open if str(path).endswith('.gz') else open | |
| with opener(path,'rt') as f: | |
| for line in f: | |
| if line.strip():yield json.loads(line) | |
| def write(path,value): | |
| p=Path(path);p.parent.mkdir(parents=True,exist_ok=True);tmp=p.with_name(p.name+'.'+str(os.getpid())+'.tmp') | |
| tmp.write_text(json.dumps(value,ensure_ascii=False,indent=2,allow_nan=False)+'\n');tmp.replace(p) | |
| def jsonl(path,values): | |
| p=Path(path);p.parent.mkdir(parents=True,exist_ok=True);tmp=p.with_name(p.name+'.'+str(os.getpid())+'.tmp') | |
| with tmp.open('w') as f: | |
| for v in values:f.write(json.dumps(v,ensure_ascii=False,separators=(',',':'),allow_nan=False)+'\n') | |
| tmp.replace(p) | |
| def position(entity,t): | |
| u=np.clip((np.asarray(t,dtype=float)-entity.get('motion_start_s',0.))/entity['motion_duration_s'],0.,1.) | |
| s=10*u**3-15*u**4+6*u**5 | |
| return np.asarray(entity['initial_xyz_m'])+np.expand_dims(s,-1)*np.asarray(entity['displacement_m']) | |
| def velocity(entity,t): | |
| raw=(np.asarray(t,dtype=float)-entity.get('motion_start_s',0.))/entity['motion_duration_s'] | |
| u=np.clip(raw,0.,1.);speed=(30*u**2-60*u**3+30*u**4)/entity['motion_duration_s'] | |
| return np.expand_dims(speed,-1)*np.asarray(entity['displacement_m']) | |
| def geometry(source,receiver,t): | |
| x,y,z=map(float,position(source,t)-position(receiver,t)) | |
| az=(-(math.degrees(math.atan2(x,-z))-receiver['yaw_deg'])+180)%360-180 | |
| return dict(doa_deg=[az,math.degrees(math.atan2(y,math.hypot(x,z)))], | |
| distance_m=math.sqrt(x*x+y*y+z*z),xyz_m=position(source,t).tolist()) | |
| def grid(scene):return list(range(0,scene['duration_samples'],STEP)) | |
| def nodes(entity,step): | |
| if not np.any(entity['displacement_m']):return [0.] | |
| start=entity.get('motion_start_s',0.);end=start+entity['motion_duration_s'] | |
| return sorted(set([0.,round(start,9),round(end,9)]+[round(float(t),9) for t in np.arange(start,end,step)])) | |
| def max_internal_gap(events): | |
| spans=sorted([pair for e in events for pair in e['activity_samples']]) | |
| if not spans:raise ValueError('No emitted target speech') | |
| merged=[] | |
| for a,b in spans: | |
| if merged and a<=merged[-1][1]:merged[-1][1]=max(merged[-1][1],b) | |
| else:merged.append([a,b]) | |
| return max(((b[0]-a[1])/SR for a,b in zip(merged,merged[1:])),default=0.) | |