Download Dataset/occluded_pattern_dataset_3000/generate_occluded_pattern_dataset.py from bilaljawaid980/GRIP-Benchmark: direct link, hf CLI and curl.
- Browser
- Download file 8.22 kB
-
https://huggingface.co/datasets/bilaljawaid980/GRIP-Benchmark/resolve/main/Dataset/occluded_pattern_dataset_3000/generate_occluded_pattern_dataset.py
- Command line
-
hf download hf://datasets/bilaljawaid980/GRIP-Benchmark/Dataset/occluded_pattern_dataset_3000/generate_occluded_pattern_dataset.py
-
curl -L -o generate_occluded_pattern_dataset.py https://huggingface.co/datasets/bilaljawaid980/GRIP-Benchmark/resolve/main/Dataset/occluded_pattern_dataset_3000/generate_occluded_pattern_dataset.py
8.22 kB
| """Generate deterministic CAPTURe-style occluded-pattern counting images.""" | |
| import argparse,json,math,random | |
| from pathlib import Path | |
| from PIL import Image,ImageDraw | |
| BG=(26,26,26);AA=3 | |
| PALETTE=((74,190,205),(239,177,82),(190,125,220),(102,201,139),(232,112,126),(118,157,230)) | |
| PATTERNS=('grid','circle','triangle');SHAPES=('dot','square','triangle');STYLES=('seamless','visible_box') | |
| def grid_pattern(rows,cols,box): | |
| x0,y0,x1,y1=box;xs=[x0+i*(x1-x0)/(cols-1) for i in range(cols)];ys=[y0+i*(y1-y0)/(rows-1) for i in range(rows)] | |
| return [(round(x,3),round(y,3)) for y in ys for x in xs] | |
| def circle_pattern(n,box): | |
| x0,y0,x1,y1=box;cx=(x0+x1)/2;cy=(y0+y1)/2;r=min(x1-x0,y1-y0)/2 | |
| return [(round(cx+r*math.cos(-math.pi/2+2*math.pi*i/n),3),round(cy+r*math.sin(-math.pi/2+2*math.pi*i/n),3)) for i in range(n)] | |
| def triangle_pattern(k,box): | |
| x0,y0,x1,y1=box;cx=(x0+x1)/2;points=[] | |
| for row in range(k): | |
| y=y0+row*(y1-y0)/(k-1);span=(x1-x0)*row/(k-1);left=cx-span/2 | |
| for col in range(row+1):points.append((round(cx if row==0 else left+col*span/row,3),round(y,3))) | |
| return points | |
| def inside(p,b):x,y=p;x0,y0,x1,y1=b;return x0<=x<=x1 and y0<=y<=y1 | |
| def distance_to_rect(p,b): | |
| x,y=p;x0,y0,x1,y1=b;dx=max(x0-x,0,x-x1);dy=max(y0-y,0,y-y1);return math.hypot(dx,dy) | |
| def valid_occluder(points,bounds,radius): | |
| hidden=[p for p in points if inside(p,bounds)];total=len(points) | |
| if len(hidden)<2 or len(hidden)/total<.15 or len(hidden)/total>.60 or total-len(hidden)<math.ceil(.4*total):return False | |
| x0,y0,x1,y1=bounds | |
| for p in points: | |
| x,y=p | |
| if inside(p,bounds): | |
| if min(x-x0,x1-x,y-y0,y1-y)<radius+2:return False | |
| elif distance_to_rect(p,bounds)<radius+2:return False | |
| return True | |
| def grid_occluder(points,rows,cols,radius,rng): | |
| xs=sorted({x for x,y in points});ys=sorted({y for x,y in points});candidates=[] | |
| for r0 in range(rows): | |
| for r1 in range(r0,rows): | |
| for c0 in range(cols): | |
| for c1 in range(c0,cols): | |
| if r0==0 and r1==rows-1 or c0==0 and c1==cols-1:continue | |
| bx0=xs[c0]-(xs[1]-xs[0])*.42;bx1=xs[c1]+(xs[1]-xs[0])*.42;by0=ys[r0]-(ys[1]-ys[0])*.42;by1=ys[r1]+(ys[1]-ys[0])*.42;b=(bx0,by0,bx1,by1) | |
| if valid_occluder(points,b,radius):candidates.append(b) | |
| if not candidates:raise RuntimeError('no valid grid occluder') | |
| return rng.choice(candidates) | |
| def general_occluder(points,pattern_box,radius,rng): | |
| x0,y0,x1,y1=pattern_box;xs=sorted({p[0] for p in points});ys=sorted({p[1] for p in points});xb=[x0-radius-4]+[(a+b)/2 for a,b in zip(xs,xs[1:])]+[x1+radius+4];yb=[y0-radius-4]+[(a+b)/2 for a,b in zip(ys,ys[1:])]+[y1+radius+4];candidates=[] | |
| for _ in range(500): | |
| xa,xbv=sorted(rng.sample(xb,2));ya,ybv=sorted(rng.sample(yb,2));b=(xa,ya,xbv,ybv) | |
| if valid_occluder(points,b,radius): | |
| visible=[p for p in points if not inside(p,b)] | |
| if max(x for x,y in visible)-min(x for x,y in visible)>.45*(x1-x0) and max(y for x,y in visible)-min(y for x,y in visible)>.45*(y1-y0):candidates.append(b) | |
| if not candidates:raise RuntimeError('no valid occluder') | |
| return rng.choice(candidates) | |
| def draw_object(draw,point,shape,radius,color): | |
| x,y=point;x*=AA;y*=AA;r=radius*AA | |
| if shape=='dot':draw.ellipse((x-r,y-r,x+r,y+r),fill=color,outline=(235,239,239),width=max(2,round(.65*AA))) | |
| elif shape=='square':draw.rectangle((x-r,y-r,x+r,y+r),fill=color,outline=(235,239,239),width=max(2,round(.65*AA))) | |
| else:draw.polygon(((x,(y-r)),(x-r*.92,y+r*.78),(x+r*.92,y+r*.78)),fill=color);draw.line(((x,y-r),(x-r*.92,y+r*.78),(x+r*.92,y+r*.78),(x,y-r)),fill=(235,239,239),width=max(2,round(.65*AA)),joint='curve') | |
| def render(path,size,positions,bounds,style,shape,radius,color): | |
| w,h=size;im=Image.new('RGB',(w*AA,h*AA),BG);draw=ImageDraw.Draw(im) | |
| for p in positions:draw_object(draw,p,shape,radius,color) | |
| x0,y0,x1,y1=bounds;rect=tuple(round(v*AA) for v in bounds) | |
| if style=='seamless':draw.rectangle(rect,fill=BG) | |
| else: | |
| draw.rectangle(rect,fill=(47,52,56),outline=(154,164,168),width=2*AA) | |
| # restrained diagonal hatch makes the occluding surface unmistakable | |
| step=14*AA | |
| start=int((x0-y1)*AA)-int((h+w)*AA) | |
| for offset in range(start,int((x1-y0)*AA)+step,step): | |
| pts=[] | |
| for yy in (y0*AA,y1*AA): | |
| xx=offset+yy | |
| if x0*AA<=xx<=x1*AA:pts.append((xx,yy)) | |
| for xx in (x0*AA,x1*AA): | |
| yy=xx-offset | |
| if y0*AA<=yy<=y1*AA:pts.append((xx,yy)) | |
| if len(pts)>=2:draw.line((pts[0],pts[1]),fill=(65,72,77),width=AA) | |
| im.resize((w,h),Image.Resampling.LANCZOS).save(path,'PNG') | |
| def questions(iid,row):return [ | |
| {'question_id':iid+'_q1','question_text':'How many objects are clearly visible (not hidden behind the occluder) in this image?','question_type':'visible_object_count','ground_truth':str(row['visible_object_count']),'answer_format':'numeric','difficulty_level':1}, | |
| {'question_id':iid+'_q2','question_text':'Assuming the pattern continues behind the occluded region, how many objects are there in total? Answer with a number in curly brackets, e.g. {12}.','question_type':'total_object_count','ground_truth':str(row['total_object_count']),'answer_format':'numeric','difficulty_level':2}, | |
| {'question_id':iid+'_q3','question_text':'What type of pattern do the objects form - a grid, a circle, or a triangular arrangement? Answer with one word.','question_type':'pattern_type','ground_truth':row['pattern_type'],'answer_format':'one_word','difficulty_level':3}, | |
| {'question_id':iid+'_q4','question_text':'How many objects are hidden behind the occluded region specifically (not counting the visible ones)? Answer with a number in curly brackets.','question_type':'occluded_object_count','ground_truth':str(row['occluded_object_count']),'answer_format':'numeric','difficulty_level':4}] | |
| def generate_one(i,images): | |
| rng=random.Random(i);size=(rng.randint(450,550),rng.randint(450,550));w,h=size;pattern=PATTERNS[(i-1)%3];shape=SHAPES[((i-1)+(i-1)//3)%3];style=STYLES[(i-1)%2];radius=rng.randint(8,11);box=(72,68,w-72,h-68) | |
| if pattern=='grid': | |
| rows=rng.randint(3,6);cols=rng.randint(3,6);positions=grid_pattern(rows,cols,box);params={'rows':rows,'cols':cols};bounds=grid_occluder(positions,rows,cols,radius,rng) | |
| elif pattern=='circle': | |
| n=rng.randint(6,14);positions=circle_pattern(n,box);params={'n_objects':n};bounds=general_occluder(positions,box,radius,rng) | |
| else: | |
| k=rng.randint(3,5);positions=triangle_pattern(k,box);params={'k':k};bounds=general_occluder(positions,box,radius,rng) | |
| flags=[inside(p,bounds) for p in positions];hidden=sum(flags);total=len(positions);iid=f'occluded_pattern_{i:04d}';color=PALETTE[(i-1)%len(PALETTE)];render(images/f'{iid}.png',size,positions,bounds,style,shape,radius,color) | |
| row={'id':iid,'image_path':f'images/{iid}.png','canvas_size':list(size),'pattern_type':pattern,'pattern_params':params,'shape_type':shape,'object_color':list(color),'object_radius_px':radius,'total_object_count':total,'visible_object_count':total-hidden,'occluded_object_count':hidden,'occluder_bounds':[round(v,3) for v in bounds],'occluder_style':style,'object_positions':[{'x':p[0],'y':p[1],'occluded':flag} for p,flag in zip(positions,flags)],'seed':i,'difficulty_score':round(.32+.4*hidden/total+.08*(pattern!='grid')+.06*(style=='seamless'),4)};row['questions']=questions(iid,row);return row | |
| def generate_dataset(n,output_dir,sample=False): | |
| root=Path(output_dir);images=root/'images';images.mkdir(parents=True,exist_ok=True);count=5 if sample else n | |
| with (root/'annotations.jsonl').open('w',encoding='utf8',newline='\n') as f: | |
| for i in range(1,count+1): | |
| f.write(json.dumps(generate_one(i,images),sort_keys=True,separators=(',',':'))+'\n') | |
| if i%250==0 or i==count:print(f'Generated {i}/{count}',flush=True) | |
| def main(): | |
| p=argparse.ArgumentParser();p.add_argument('--n',type=int,default=3000);p.add_argument('--output-dir',default='.');p.add_argument('--sample',action='store_true');a=p.parse_args();generate_dataset(a.n,a.output_dir,a.sample) | |
| if __name__=='__main__':main() | |