File size: 15,569 Bytes
c3f0e3f | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 | {
"schema": "efe-task-dag/1",
"taxonomy": "v2",
"taskKey": "robopaint/acrylic_vision",
"suite": "robopaint",
"base": "robopaint-acrylic-vision-i00",
"title_zh": "丙烯临摹讲道后的幻象",
"summary_zh": "将目标图临摹到主画纸,使颜色、结构和漆区覆盖同时达到图像阈值,且不超过390笔、32300毫米带漆路径及0.1的绕填路径占比。",
"embodiment_zh": "两种模式均为固定底座的七自由度Franka Panda单臂,末端为安装工具:刚性笔杆及15毫米软刷毛的圆头丙烯笔刷,替代夹爪、无开合手指;privileged可重置与查询画纸和评分,standard仅通过相机及本体反馈控制同一画笔。",
"task_class": "compositional",
"analysis_level": "step",
"spec": {
"version": "1.2",
"task": "robopaint-acrylic-vision-i00",
"bench": "robopaint",
"progress": "live",
"entities": {},
"milestones": [
{
"id": "brush_loaded",
"zh": "笔刷蘸颜料",
"mode": "event",
"when": null,
"after": [],
"hold_s": 0,
"goal": false,
"criterion_zh": "场景视角中,笔尖降入主纸旁双排颜料孔的任意一孔,刷毛随后带色或能在纸上留下有色笔迹;对应加载条件为笔尖到该孔中心的平面距离小于12毫米且世界高度低于0.028米。仅在调色盘上方经过、未进入颜料面且仍不能留下色迹时未完成,首次成功加载后本步骤保持已完成。",
"step": {
"cat": "其他",
"item": "蘸颜料"
},
"contact": [
{
"who": "手持工具",
"how": "其他",
"note_zh": "单臂末端固定圆头笔刷的刷毛浸入颜料孔中的颜料,候选接触为浸蘸;液体加载不属于抓握、压擦或从下方托舀,词表暂无合适条目。",
"plan": 1
}
],
"optional": false,
"verify_zh": null,
"sources": [
"inputs/task/privileged/instruction.md:10-15",
"inputs/task/standard/instruction.md:26-32",
"inputs/grader/efe_robopaint-files/rcb_brush.py:432-438 干笔复位",
"inputs/grader/efe_robopaint-files/rcb_brush.py:274-279 色孔内且低于液面",
"inputs/grader/efe_robopaint-files/rcb_brush.py:875-885 加载颜料",
"inputs/task/privileged/tests/instance.json:59-84 颜料面与首个色孔",
"inputs/oracle/replay.mp4 @ 0.0-0.4 s 干净画纸至首个有色笔迹",
"inputs/grader/efe_robopaint-files/rcb_brush.py:1479-1493 固定底座与替代夹爪的安装笔刷"
],
"conf": 7,
"note_zh": "干笔初态强制先加载颜料,按DECISIONS.md D11当前行为保留此前提;不要求特定颜色先蘸,也不把每次换色拆成步骤。蘸颜料是笔刷的材料加载状态,现有画/写和舀均不准确,使用其他并保留类型缺口;笔刷已固定安装,不建拿起笔刷节点。"
},
{
"id": "painting_copied",
"zh": "临摹目标画",
"mode": "state",
"when": null,
"after": [
"brush_loaded"
],
"hold_s": 0,
"freeze_held": false,
"goal": true,
"criterion_zh": "主纸俯视图(成功录像右下格)呈现目标的红橙底色、左上至右中的棕色斜带、下部浅灰白弧形与黑色块及右上黑色形状,整体颜色、位置和笔触接近目标;验收在每轴最多5毫米的整像素平移配准后进行,最优移位落在搜索边界则不用平移,两图以1毫米标准差模糊,任一图有漆的区域外扩2毫米后平均CIELAB色差不大于2、色差90百分位不大于6,亮度结构相似度不小于0.92(窗口标准差1.5毫米)、边缘相关不小于0.8、覆盖率至少0.5的漆区交并比不小于0.95,且主纸带漆接触不超过390笔、带漆压纸总路径不超过32300毫米、长度除以至少10毫米的包围盒对角线大于2.5的笔画所占路径比例不超过0.1。空白纸、仅画出部分底色或主要形状缺失均未完成,后续覆盖导致任一验收条件不成立时失去完成状态;无需松笔、退开、回初位或持续静置。",
"step": {
"cat": "材料",
"item": "画/写"
},
"contact": [
{
"who": "手持工具",
"how": "压/擦",
"note_zh": "单臂末端固定圆头丙烯笔刷的带漆刷毛压住主画纸并扫过纸面,按压深度改变笔触宽度。",
"plan": 1
}
],
"optional": false,
"verify_zh": null,
"sources": [
"inputs/task/privileged/instruction.md:26-53",
"inputs/task/standard/instruction.md:40-65",
"inputs/task/privileged/tests/instance.json:215-230 固定成功阈值",
"inputs/grader/efe_robopaint-files/rcb_brush_score.py:365-395 全纸平移配准",
"inputs/grader/efe_robopaint-files/rcb_brush_score.py:398-458 图像指标与成功条件",
"inputs/grader/efe_robopaint-files/rcb_brush.py:1323-1332 未抬笔的当前笔画也计分",
"inputs/grader/efe_robopaint-files/rcb_brush_strict.py:74-91 路径与包围盒",
"inputs/grader/efe_robopaint-files/rcb_brush_strict.py:140-193 过程规则及联合成功",
"inputs/oracle/replay.mp4 @ 5.0,15.0,25.0,35.0,45.0,55.0,58.6 s 成画过程及终态",
"inputs/oracle/run.json:reward 成功时各项指标",
"inputs/grader/efe_robopaint-files/rcb_brush_score.py:332-350 目标图及原始覆盖率掩模"
],
"conf": 6,
"note_zh": "整幅图是一个语义目标;隐藏的300笔四层程序只是示范方案,390是上限而非要求重复390次,不按笔数、颜色、层或图像指标建节点。过程规则是本步验收;笔数和路径一旦超限不可撤销,另列历史否决约束;涂画路径占比可随新增合规笔画下降,只在终局验收,不作历史否决。结果在笔刷仍接触纸面时也可成立,故freeze_held为false。"
}
],
"constraints": [
{
"id": "stroke_budget_exceeded",
"zh": "超过笔数上限",
"never": null,
"criterion_zh": "主画纸累计带漆连续接触超过390笔;练习纸和干笔接触不计,仍压在纸上的当前带漆笔画也计入。一旦超过上限,本回合不能通过后续加画恢复。",
"consequence": "历史否决",
"sources": [
"inputs/task/privileged/instruction.md:4",
"inputs/task/privileged/tests/instance.json:226",
"inputs/grader/efe_robopaint-files/rcb_brush_score.py:440-456",
"inputs/grader/efe_robopaint-files/rcb_brush.py:1329-1332"
]
},
{
"id": "paint_path_budget_exceeded",
"zh": "超过画迹路径上限",
"never": null,
"criterion_zh": "主画纸所有带漆压纸笔画的笔尖平面路径长度之和超过32300毫米,按100赫兹物理采样累加;练习纸、空中移笔和干笔不计。累计长度不可减少,超限后该回合被否决。",
"consequence": "历史否决",
"sources": [
"inputs/task/privileged/instruction.md:42-44",
"inputs/task/privileged/tests/instance.json:227",
"inputs/grader/efe_robopaint-files/rcb_brush_strict.py:74-83",
"inputs/grader/efe_robopaint-files/rcb_brush_strict.py:145-160"
]
}
],
"success": null,
"atoms": [
{
"atom": "brush_loaded",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush.py:960 MyPaintPainter.brush_state()[loaded]; log dyn.brush_loaded per frame"
},
{
"atom": "de_mean",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[de_mean]; log dyn.de_mean per frame; no state logs provided"
},
{
"atom": "de_p90",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[de_p90]; log dyn.de_p90 per frame; no state logs provided"
},
{
"atom": "ssim_L",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[ssim_L]; log dyn.ssim_L per frame; no state logs provided"
},
{
"atom": "edge_corr",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[edge_corr]; log dyn.edge_corr per frame; no state logs provided"
},
{
"atom": "paint_iou",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[paint_iou]; log dyn.paint_iou per frame; no state logs provided"
},
{
"atom": "n_strokes",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[n_strokes]; log dyn.n_strokes per frame; no state logs provided"
},
{
"atom": "path_mm",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[path_mm]; log dyn.path_mm per frame; no state logs provided"
},
{
"atom": "scribble_share",
"level": "new_code",
"source": "inputs/grader/efe_robopaint-files/rcb_brush_strict.py:209-232 Session.score()[scribble_share]; log dyn.scribble_share per frame; no state logs provided"
}
]
},
"native_metrics": [
{
"name": "de_mean",
"unit": "DeltaE_ab",
"threshold": 2,
"direction": "lower_better"
},
{
"name": "de_p90",
"unit": "DeltaE_ab",
"threshold": 6,
"direction": "lower_better"
},
{
"name": "ssim_L",
"unit": "1",
"threshold": 0.92,
"direction": "higher_better"
},
{
"name": "edge_corr",
"unit": "1",
"threshold": 0.8,
"direction": "higher_better"
},
{
"name": "paint_iou",
"unit": "1",
"threshold": 0.95,
"direction": "higher_better"
},
{
"name": "n_strokes",
"unit": "stroke",
"threshold": 390,
"direction": "lower_better"
},
{
"name": "path_mm",
"unit": "mm",
"threshold": 32300,
"direction": "lower_better"
},
{
"name": "scribble_share",
"unit": "1",
"threshold": 0.1,
"direction": "lower_better"
},
{
"name": "de_mean_coarse",
"unit": "DeltaE_ab",
"threshold": null,
"direction": "lower_better"
},
{
"name": "ssim_L_fine",
"unit": "1",
"threshold": null,
"direction": "higher_better"
},
{
"name": "region_cov_min",
"unit": "1",
"threshold": null,
"direction": "higher_better"
},
{
"name": "region_cov_mean",
"unit": "1",
"threshold": null,
"direction": "higher_better"
}
],
"open_questions_zh": [
"inputs/states/index.json的logs为空,无逐帧状态日志;所有节点when、约束never及success均为null,不能把run.json的终局reward当成逐帧状态字段。需新增记录笔刷加载和原生绘画指标,才能自动核验。视频能判断主要色块和未完成状态,不能单独精确确认接近阈值的色差或累计路径。",
"按DECISIONS.md D11保留干笔首次蘸料这一必要前提,蘸料与成画可暂停分开,故从文字及明确初态得到两个步骤并依P17判为compositional;保留或省略蘸料均已获允许。蘸颜料及其浸蘸接触无准确词表项,分别使用其他·蘸颜料和手持工具×其他,不将吸附颜料解释为舀。",
"按D6与D11将整幅图作为一个临摹步骤;隐藏笔画程序、四个绘制层及每种颜色不是文本要求的独立结果,390笔是上限,不是计数目标。",
"模式差异:共同物理验收均为is_success_paint与strict_ok。privileged评分器还要求合法轨迹在默认两个独立进程重放均成功、最终状态哈希相同且固定实例匹配(inputs/grader/efe_robopaint-files/privileged/grade.py:89-129);standard评分实际回合终态,并检查已终结、有live判决、无基础设施错误和模拟器错误(standard/grade.py:29-49,sim/server.py:86-118)。这些交付及重放有效性条件没有可绑定状态日志,记录在此,不增加语义步骤,success保留null。两次重放不是独立任务场景。",
"文本要求保持笔刷竖直,但评分器只记录max_tilt_deg_in_contact而没有成功或失败阈值;按S13不把倾角、按压深度、接触速度、蘸料次数或操作成本记录建成约束。de_mean_coarse、ssim_L_fine及region_cov_min/mean也只是报告值,本实例target.success未启用这些附加阈值(rcb_brush_score.py:413-458)。",
"scribble_share以终局总路径为分母,中途高于0.1仍可能通过新增合规笔画降回阈值,故只作为画面目标的终局验收;不将任何一次中途超比例误标为历史否决。笔数和总路径上限单调累加,超限才作为历史否决约束。",
"唯一成功示范为inputs/oracle/oracle.json所选tvzr6sk,亦是后续待标轨迹(is_labelled_run=true),无独立成功或状态日志可交叉验证;只用其画面与终局指标确认目标外观和可行性,不从其颜色顺序、385笔或触笔方案增加边。",
"漆区交并比使用目标truth_alpha.npz中的原始覆盖率掩模(rcb_brush_score.py:332-350),不是直接从目标图颜色估出的掩模;本次输入未提供该二进制真值,不能仅凭截图重算精确交并比。"
],
"conf": 6,
"conf_reason_zh": "文字、目标图、固定初态和两模式评分代码支持蘸料后完成整幅临摹,阈值均有来源,成功录像及终局指标相符;但没有逐帧状态日志,全部条件需视频判读,精确图像及路径阈值无法自动核验,蘸料类型仍有词表缺口。",
"review": {
"status": "agreed",
"rounds": 1,
"reviewers": [
{
"engine": "claude",
"model": "claude-sonnet-5-5",
"round": 1,
"verdict": "agree",
"conf": 7,
"summary_zh": "文本只描述临摹整幅画(含390笔上限与过程规则);干笔蘸料按D11可保留,两节点、边brush_loaded→painting_copied物理必要,整幅图为一个步骤,符合D6/D11。成功阈值(ΔE≤2、p90≤6、SSIM≥0.92、边缘≥0.8、IoU≥0.95、390笔、路径≤32300、scribble_share≤0.1)与指令及评分器一致;无状态日志故when/never/success为null并已在open_questions_zh说明。笔数与路径作为历史否决约束,同时写在验收文字内,属可接受的重复,不影响节点集合。无需修改。",
"n_issues": 0,
"n_major": 0
}
],
"open_zh": []
},
"evidence_check": {
"n_state_runs": 0,
"n_faithful": 0,
"success_runs_reaching_all": "0/0",
"spec_success_eq_grader": "0/0",
"diverged_runs": [],
"notes_zh": [
"when=null 的节点不在状态日志上检查:['brush_loaded', 'painting_copied']"
]
},
"provenance": {
"prompt": "prompts/01_task_dag.md",
"prompt_sha256": "1142cdb56758062bbd0ff585ad99f3fefe07a96377b61614dc4f50b2d16600b7",
"engine": "codex",
"model": "gpt-6.1-sol",
"effort": "high",
"created": "2026-10-29T01:47:37Z",
"review_prompt": "prompts/02_task_dag_review.md",
"review_prompt_sha256": "65812ffc3cb33aaeb10dbe28c925ffe5535a356f8132ec431dcf58b59c2a0b85",
"reconcile_prompt": "prompts/05_reconcile.md",
"reconcile_prompt_sha256": "b3cd3803cf22f21db996e7ab1145811f823e92c7728c6f21d3b6649cc1ed0e83",
"calls": [
{
"stage": "dag",
"round": 0,
"ok": true,
"space": "work/tasks/robopaint__acrylic_vision/01_dag"
},
{
"stage": "dag_review",
"round": 1,
"ok": true,
"space": "work/tasks/robopaint__acrylic_vision/02_review_r1"
}
],
"review_engine": "claude",
"review_model": "claude-sonnet-5-5",
"inputs_manifest_sha256": "643edf63c090ac28c9fd1c01210c07b7bd9102919daae163be3bb77a90e02800",
"repo_commit": "4c6d01b9990e3b78ddb620bbd237932b3696f60e"
}
}
|