AEA 单人序列(script 1 / 4 / 5):只用 ASR + sound 事件做 goal / sub-goal(2026-10-27)
范围:本地 AEA 单人 script(1/4/5)共 16 条录像(loc2_script1_seq4 下载为空,未计入);只保留模型判断 others_involved = false 的 11 条纯单人独立完成的录像(剔除 loc2_script1_seq2 = true,loc1_script1_seq1 / loc5_script4_seq6 = unclear,loc5_script4_seq1/2 事件太少无法判断)。事件 = official speech.csv ASR 句子(不分说话人,可能是电视 / 自言自语 / 呼吸声)+ ours-legacy sound 声音标签(未核验)。SAVVY-Bench 没有标注单人录像,所以没有说话人 / 方向信息。goal 草稿 = Gemini 3.1 Pro,activity_type 取自官方 script 活动列表,sub-goal 只引用事件 id,时间由代码复制;未人工核验。others_involved 是模型对「是否有他人参与」的判断,用来筛纯单人独立完成的录像。narration 不使用。11 条全部有草稿。
Gallery
loc1_script1_seq3_rec1[cleaning] Vacuum the room and prepare party decorations
Script 1 · lazy morning before a party · 377 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helps
Script 1 · lazy morning before a party · 377 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helps
loc2_script1_seq1_rec1[reading, games or exercise] Enter the room, turn on the TV, and use a phone to relax.
Script 1 · lazy morning before a party · 204 s · others_involved: False
audio_helpsaudio_helpsaudio_helps
Script 1 · lazy morning before a party · 204 s · others_involved: False
audio_helpsaudio_helpsaudio_helps
loc3_script1_seq1_rec1[reading, games or exercise] Operate a remote control and type on a smartphone
Script 1 · lazy morning before a party · 151 s · others_involved: False
audio_helpsaudio_helps
Script 1 · lazy morning before a party · 151 s · others_involved: False
audio_helpsaudio_helps
loc3_script1_seq2_rec1[reading, games or exercise] Browse TV and radio channels to find a morning sports game.
Script 1 · lazy morning before a party · 228 s · others_involved: False
audio_helpsspeech_onlysound_only
Script 1 · lazy morning before a party · 228 s · others_involved: False
audio_helpsspeech_onlysound_only
loc3_script4_seq2_rec1[making coffee] Brew coffee using a French press and a stovetop kettle.
Script 4 · relaxed Sunday morning · 312 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helps
Script 4 · relaxed Sunday morning · 312 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helpsaudio_helps
loc3_script4_seq7_rec1[reading, games or exercise] Play a game on a mobile phone.
Script 4 · relaxed Sunday morning · 169 s · others_involved: False
sound_onlysound_onlysound_only
Script 4 · relaxed Sunday morning · 169 s · others_involved: False
sound_onlysound_onlysound_only
loc3_script5_seq3_rec1[indoor-outdoor transition] Remove outerwear or change clothes and check phone
Script 5 · getting home and getting going · 113 s · others_involved: False
audio_helpsaudio_helps
Script 5 · getting home and getting going · 113 s · others_involved: False
audio_helpsaudio_helps
loc3_script5_seq6_rec1[indoor-outdoor transition] Enter the home, organize clothing into a bag, and check a smartphone.
Script 5 · getting home and getting going · 158 s · others_involved: False
audio_helpsaudio_helpsaudio_helps
Script 5 · getting home and getting going · 158 s · others_involved: False
audio_helpsaudio_helpsaudio_helps
loc4_script1_seq1_rec1[reading, games or exercise] Perform an exercise routine
Script 1 · lazy morning before a party · 165 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helps
Script 1 · lazy morning before a party · 165 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helps
loc4_script1_seq3_rec1[cleaning] Vacuum the room
Script 1 · lazy morning before a party · 109 s · others_involved: False
sound_onlyaudio_helps
Script 1 · lazy morning before a party · 109 s · others_involved: False
sound_onlyaudio_helps
loc4_script5_seq3_rec1[organization and laundry] Organize clothes, handle a suitcase, and read a book
Script 5 · getting home and getting going · 159 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helps
Script 5 · getting home and getting going · 159 s · others_involved: False
audio_helpsaudio_helpsaudio_helpsaudio_helps
loc1_script1_seq3_rec1
Script 1 · lazy morning before a party · 377 s · ASR utterances 1 · legacy sound tags 31
video: SAVVY Video/loc1_script1_seq3_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
g4
g5
g6
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:cleaning
overall_goal:Vacuum the room and prepare party decorations
others_involved(模型判断有无他人参与):False
unrelated events:X013 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Vacuum the first area | 0:00.1–0:31.4 | mixed | audio_helps | S000 X000 X001 | The wearer announces their cleaning duties in S000 and begins operating the vacuum cleaner in X000 and X001. |
| g2 | Relocate and plug in the vacuum cleaner | 0:30.4–0:57.8 | sound | audio_helps | X002 X003 X004 X005 X006 | Sounds of unplugging, handling the device, and plugging it back in from X002 to X006 indicate moving the vacuum to a new outlet. |
| g3 | Continue vacuuming the room | 1:05.1–2:54.2 | sound | audio_helps | X007 X008 X009 X010 X011 X012 | The vacuum cleaner powers up and operates again from X007 to X012, with a zipper sound in X010 possibly indicating adjusting the vacuum bag or moving an obstacle. |
| g4 | Unpack or handle paper materials | 3:07.0–4:01.7 | sound | audio_helps | X014 X015 X016 X017 X018 X019 | Continuous rustling of paper from X014 to X019 suggests the wearer is unpacking items or preparing paper decorations. |
| g5 | Place objects on surfaces | 4:00.3–4:11.1 | sound | audio_helps | X020 X021 X022 X023 X024 | Repeated sounds of objects touching a surface from X020 to X024 indicate the wearer is setting down the items they just unpacked. |
| g6 | Handle party balloons | 4:36.7–5:46.1 | sound | audio_helps | X025 X026 X027 X028 X029 X030 | Balloon squeaks from X025 to X030 show the wearer is inflating, tying, or arranging balloons for the upcoming party. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:00.1–0:01.8 | speech | as part of my cleaning duties. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:28.9–0:29.7 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:30.2–0:31.4 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:30.4–0:30.4 | sound | Unplug [appliance] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:39.7–0:47.8 | sound | Handling device | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:47.8–0:47.8 | sound | Unplug [appliance] | ours-legacy sound (unverified, model/hand made) |
| X005 | 0:53.2–0:57.8 | sound | Handling device | ours-legacy sound (unverified, model/hand made) |
| X006 | 0:55.5–0:56.5 | sound | Plugging in | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:05.1–1:29.3 | sound | Vacuum cleaner [Vacuum] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:36.7–1:59.0 | sound | Vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X009 | 2:30.2–2:30.6 | sound | Vacuum cleaner power up [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X010 | 2:30.2–2:32.6 | sound | zipper [bag] | ours-legacy sound (unverified, model/hand made) |
| X011 | 2:37.3–2:42.2 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X012 | 2:52.5–2:54.2 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X013 | 3:00.0–3:30.0 | sound | Music [background] | ours-legacy sound (unverified, model/hand made) |
| X014 | 3:07.0–3:09.0 | sound | Rustling | ours-legacy sound (unverified, model/hand made) |
| X015 | 3:13.1–3:15.4 | sound | Rustling | ours-legacy sound (unverified, model/hand made) |
| X016 | 3:27.5–3:30.1 | sound | Rustling | ours-legacy sound (unverified, model/hand made) |
| X017 | 3:30.2–3:31.1 | sound | Paper rustling [paper] | ours-legacy sound (unverified, model/hand made) |
| X018 | 3:35.6–3:37.8 | sound | Paper rustling [paper] | ours-legacy sound (unverified, model/hand made) |
| X019 | 3:52.9–4:01.7 | sound | Paper rustling [paper] | ours-legacy sound (unverified, model/hand made) |
| X020 | 4:00.3–4:00.8 | sound | Object touching surface [Object] | ours-legacy sound (unverified, model/hand made) |
| X021 | 4:01.2–4:01.8 | sound | Object touching surface [Object] | ours-legacy sound (unverified, model/hand made) |
| X022 | 4:01.8–4:02.2 | sound | Object touching surface [Object] | ours-legacy sound (unverified, model/hand made) |
| X023 | 4:09.1–4:09.4 | sound | Object touching surface [Object] | ours-legacy sound (unverified, model/hand made) |
| X024 | 4:10.8–4:11.1 | sound | Object touching surface [Object] | ours-legacy sound (unverified, model/hand made) |
| X025 | 4:36.7–4:36.7 | sound | Balloon squeak [Balloon] | ours-legacy sound (unverified, model/hand made) |
| X026 | 4:38.6–4:38.6 | sound | Balloon squeak [Balloon] | ours-legacy sound (unverified, model/hand made) |
| X027 | 5:30.8–5:31.0 | sound | balloon squeak [balloon] | ours-legacy sound (unverified, model/hand made) |
| X028 | 5:34.5–5:34.8 | sound | balloon squeak [balloon] | ours-legacy sound (unverified, model/hand made) |
| X029 | 5:41.9–5:42.2 | sound | balloon squeak [balloon] | ours-legacy sound (unverified, model/hand made) |
| X030 | 5:45.6–5:46.1 | sound | balloon squeak [balloon] | ours-legacy sound (unverified, model/hand made) |
loc2_script1_seq1_rec1
Script 1 · lazy morning before a party · 204 s · ASR utterances 7 · legacy sound tags 9
video: SAVVY Video/loc2_script1_seq1_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:reading, games or exercise
overall_goal:Enter the room, turn on the TV, and use a phone to relax.
others_involved(模型判断有无他人参与):False
unrelated events:S000, S001, S002, S003 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Open the door and enter the room. | 1:15.0–1:26.6 | sound | audio_helps | X000 X001 | Sound tags for creaking wood and a door opening in X000 and X001 indicate the wearer entering the room. |
| g2 | Turn on and tune the television. | 1:30.2–2:59.1 | mixed | audio_helps | X002 S004 X003 S005 X004 S006 X005 X006 | The wearer uses a remote control to turn on the TV in X002 and X003, reacts to the screen in S004 and S005, and clicks again in X004 until settling on a channel in S006, X005, and X006. |
| g3 | Type on a mobile phone. | 3:05.7–3:24.2 | sound | audio_helps | X007 X008 | Typing sounds in X007 and X008 indicate the wearer is using their phone. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:19.8–0:20.8 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| S001 | 0:42.9–0:43.9 | speech | Okay. | official speech.csv (ASR, speaker unknown) |
| S002 | 0:48.1–0:49.1 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| S003 | 1:11.7–1:12.7 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X000 | 1:15.0–1:15.5 | sound | creak [wood] | ours-legacy sound (unverified, model/hand made) |
| X001 | 1:26.3–1:26.6 | sound | Door opens. [door] | ours-legacy sound (unverified, model/hand made) |
| X002 | 1:30.2–1:31.1 | sound | Remote control clicks [remote control] | ours-legacy sound (unverified, model/hand made) |
| S004 | 1:30.9–1:31.9 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X003 | 1:41.8–1:42.1 | sound | TV turns on [TV] | ours-legacy sound (unverified, model/hand made) |
| S005 | 2:00.8–2:01.8 | speech | Interesting. | official speech.csv (ASR, speaker unknown) |
| X004 | 2:02.3–2:21.6 | sound | remote control, clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| S006 | 2:19.5–2:21.4 | speech | Oh there we go. | official speech.csv (ASR, speaker unknown) |
| X005 | 2:32.1–2:36.8 | sound | TV static [TV] | ours-legacy sound (unverified, model/hand made) |
| X006 | 2:36.8–2:59.1 | sound | TV [TV] | ours-legacy sound (unverified, model/hand made) |
| X007 | 3:05.7–3:14.3 | sound | Typing [Phone] | ours-legacy sound (unverified, model/hand made) |
| X008 | 3:22.9–3:24.2 | sound | Typing [Phone] | ours-legacy sound (unverified, model/hand made) |
loc3_script1_seq1_rec1
Script 1 · lazy morning before a party · 151 s · ASR utterances 11 · legacy sound tags 22
video: SAVVY Video/loc3_script1_seq1_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:reading, games or exercise
overall_goal:Operate a remote control and type on a smartphone
others_involved(模型判断有无他人参与):False
unrelated events:X000, S000, X001, S001, X002, S002, S003, X003, X004, S004, X005, X006, S005, S006, S007 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Operate a device using a remote control | 1:30.5–2:17.3 | mixed | audio_helps | X007 X008 X009 X010 X011 X012 X013 X014 X015 X016 X017 S008 X018 X019 S009 | The wearer clicks a remote control multiple times, occasionally saying 'Alright' to confirm actions. |
| g2 | Walk to a new location and type on a smartphone | 2:18.3–2:29.7 | mixed | audio_helps | X020 S010 X021 | The wearer walks and then types on a smartphone, uttering a brief sound. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| X000 | 0:02.0–0:04.0 | sound | breath [person] | ours-legacy sound (unverified, model/hand made) |
| S000 | 0:10.6–0:11.6 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X001 | 0:12.7–0:24.1 | sound | breath [person] | ours-legacy sound (unverified, model/hand made) |
| S001 | 0:22.5–0:24.5 | speech | Starting a recording. | official speech.csv (ASR, speaker unknown) |
| X002 | 0:25.5–0:29.2 | sound | breath [person] | ours-legacy sound (unverified, model/hand made) |
| S002 | 0:27.1–0:28.1 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| S003 | 0:43.9–0:45.7 | speech | Alright. Uh | official speech.csv (ASR, speaker unknown) |
| X003 | 0:47.2–0:48.4 | sound | Telephone ringing [telephone] | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:48.6–0:50.4 | sound | Telephone ringing [telephone] | ours-legacy sound (unverified, model/hand made) |
| S004 | 0:48.8–0:51.8 | speech | we're recording. Hold on. | official speech.csv (ASR, speaker unknown) |
| X005 | 0:54.0–0:55.7 | sound | Telephone ringing [telephone] | ours-legacy sound (unverified, model/hand made) |
| X006 | 0:57.6–0:59.3 | sound | Telephone ringing [telephone] | ours-legacy sound (unverified, model/hand made) |
| S005 | 1:05.5–1:06.5 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| S006 | 1:11.4–1:12.4 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| S007 | 1:19.7–1:21.5 | speech | Mm hmm. | official speech.csv (ASR, speaker unknown) |
| X007 | 1:30.5–1:31.0 | sound | Remote control clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:38.0–1:38.7 | sound | Remote control clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| X009 | 1:43.7–1:44.0 | sound | Remote control clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| X010 | 1:48.4–1:49.0 | sound | Remote control clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| X011 | 1:56.3–1:56.6 | sound | Remote control clicking [remote control] | ours-legacy sound (unverified, model/hand made) |
| X012 | 2:02.4–2:02.5 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X013 | 2:03.7–2:03.8 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X014 | 2:04.8–2:05.0 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X015 | 2:06.8–2:07.0 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X016 | 2:07.8–2:08.0 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X017 | 2:08.8–2:09.0 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| S008 | 2:10.0–2:11.3 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X018 | 2:11.5–2:11.6 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| X019 | 2:13.7–2:13.8 | sound | Click [remote control] | ours-legacy sound (unverified, model/hand made) |
| S009 | 2:16.3–2:17.3 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X020 | 2:18.3–2:18.8 | sound | Walk [person] | ours-legacy sound (unverified, model/hand made) |
| S010 | 2:18.8–2:19.8 | speech | Uh. | official speech.csv (ASR, speaker unknown) |
| X021 | 2:28.2–2:29.7 | sound | Typing [smartphone] | ours-legacy sound (unverified, model/hand made) |
loc3_script1_seq2_rec1
Script 1 · lazy morning before a party · 228 s · ASR utterances 17 · legacy sound tags 19
video: SAVVY Video/loc3_script1_seq2_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:reading, games or exercise
overall_goal:Browse TV and radio channels to find a morning sports game.
others_involved(模型判断有无他人参与):False
unrelated events:S001, S004, S005, S006, S009, S010, S011, S012, S013, S014, S016 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Browse TV channels to find a specific morning game. | 0:02.9–1:46.3 | mixed | audio_helps | S000 X000 X001 X002 X003 X004 X005 S002 X006 S003 X007 X008 X009 | The wearer clicks the remote multiple times and mentions looking for a morning game. |
| g2 | Troubleshoot the media setup by switching to a different remote or machine. | 1:59.0–2:15.8 | speech | speech_only | S007 S008 | The wearer realizes they need to use a different machine or remote to get the desired channel. |
| g3 | Tune the radio or TV to find the sports broadcast. | 2:30.0–3:40.1 | sound | sound_only | X010 X011 X012 X013 X014 X015 S015 X016 X017 X018 | Sounds of radio tuning, TV static, and remote clicks indicate the wearer is actively searching for the right station. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:02.9–0:03.9 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:20.7–0:21.0 | sound | Click [TV remote] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:35.2–0:35.9 | sound | remote control click [remote] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:38.3–0:39.0 | sound | remote control click [remote] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:41.4–0:42.0 | sound | remote control click [remote] | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:44.6–0:45.3 | sound | remote control click [remote] | ours-legacy sound (unverified, model/hand made) |
| S001 | 0:46.1–0:55.5 | speech | Just stick around. I probably should be happy. We'll redeem ourselves with a giant another player but it's going to be a giant player. George Contos. | official speech.csv (ASR, speaker unknown) |
| X005 | 0:53.6–0:54.1 | sound | remote control click [remote] | ours-legacy sound (unverified, model/hand made) |
| S002 | 0:56.6–0:57.6 | speech | Hmm. | official speech.csv (ASR, speaker unknown) |
| X006 | 1:02.3–1:02.8 | sound | Remote control click [Remote control] | ours-legacy sound (unverified, model/hand made) |
| S003 | 1:06.2–1:10.9 | speech | It's a morning game. It's supposed to be here. | official speech.csv (ASR, speaker unknown) |
| S004 | 1:12.0–1:14.3 | speech | I don't know. It's a rebound. | official speech.csv (ASR, speaker unknown) |
| X007 | 1:21.8–1:22.4 | sound | Remote control click [Remote control] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:26.1–1:26.6 | sound | Remote control click [Remote control] | ours-legacy sound (unverified, model/hand made) |
| X009 | 1:45.0–1:46.3 | sound | TV channel change sound effect [TV] | ours-legacy sound (unverified, model/hand made) |
| S005 | 1:47.0–1:48.7 | speech | Starting to start doing | official speech.csv (ASR, speaker unknown) |
| S006 | 1:50.8–1:56.8 | speech | something else. Yeah, regular science. | official speech.csv (ASR, speaker unknown) |
| S007 | 1:59.0–2:12.2 | speech | Oh. Maybe I need to use one of these ones. Let's see. Using the machine. I don't use it. I use the machine. Oh, yeah. There we go. Okay. | official speech.csv (ASR, speaker unknown) |
| S008 | 2:14.8–2:15.8 | speech | Uh. | official speech.csv (ASR, speaker unknown) |
| S009 | 2:17.5–2:18.5 | speech | The | official speech.csv (ASR, speaker unknown) |
| S010 | 2:21.1–2:22.1 | speech | seventh | official speech.csv (ASR, speaker unknown) |
| X010 | 2:30.0–2:31.4 | sound | Radio tuning noise [radio] | ours-legacy sound (unverified, model/hand made) |
| S011 | 2:32.1–2:34.5 | speech | round last night in the Atlanta | official speech.csv (ASR, speaker unknown) |
| X011 | 2:37.1–2:39.7 | sound | Radio tuning noise [radio] | ours-legacy sound (unverified, model/hand made) |
| S012 | 2:39.7–2:40.7 | speech | Draft. | official speech.csv (ASR, speaker unknown) |
| S013 | 2:44.4–2:47.5 | speech | Second overall by Indiana. Hmm. | official speech.csv (ASR, speaker unknown) |
| S014 | 2:52.2–3:13.7 | speech | Well, you got your wish but he's not going to stop. Yeah, all the sports. | official speech.csv (ASR, speaker unknown) |
| X012 | 2:57.2–2:58.9 | sound | Radio tuning noise [radio] | ours-legacy sound (unverified, model/hand made) |
| X013 | 3:00.0–3:04.1 | sound | TV static [TV] | ours-legacy sound (unverified, model/hand made) |
| X014 | 3:04.1–3:04.9 | sound | Remote control clicking [Remote control] | ours-legacy sound (unverified, model/hand made) |
| X015 | 3:08.5–3:15.8 | sound | TV static [TV] | ours-legacy sound (unverified, model/hand made) |
| S015 | 3:18.3–3:19.3 | speech | Hmm. | official speech.csv (ASR, speaker unknown) |
| X016 | 3:19.3–3:21.4 | sound | Remote control clicking [Remote control] | ours-legacy sound (unverified, model/hand made) |
| X017 | 3:22.1–3:29.9 | sound | TV static [TV] | ours-legacy sound (unverified, model/hand made) |
| S016 | 3:24.7–3:25.7 | speech | Everything | official speech.csv (ASR, speaker unknown) |
| X018 | 3:31.2–3:40.1 | sound | TV [television] | ours-legacy sound (unverified, model/hand made) |
loc3_script4_seq2_rec1
Script 4 · relaxed Sunday morning · 312 s · ASR utterances 1 · legacy sound tags 21
video: SAVVY Video/loc3_script4_seq2_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
g4
g5
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:making coffee
overall_goal:Brew coffee using a French press and a stovetop kettle.
others_involved(模型判断有无他人参与):False
unrelated events:X020 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Start boiling water on the stove. | 0:08.0–1:28.9 | mixed | audio_helps | S000 X000 X001 X002 | The wearer says a brief confirmation word to begin the task, followed by the sound of a gas burner igniting and water bubbling in the kettle. |
| g2 | Prepare the coffee grounds and French press. | 1:28.9–2:07.2 | sound | audio_helps | X003 X004 X005 X006 | Sounds of a coffee filter, French press assembly, and opening a coffee bag indicate setting up the brewing equipment and grounds. |
| g3 | Adjust the kettle and reignite the stove. | 2:16.3–3:28.4 | sound | audio_helps | X007 X008 X009 X010 X011 X012 | The wearer handles the kettle lid, places the kettle back on the stovetop, and ignites the gas stove again to heat more water. |
| g4 | Pour hot water into the coffee container. | 3:36.9–3:59.7 | sound | audio_helps | X013 X014 X015 X016 | The wearer pours water from the kettle to fill a container, then places the kettle down and covers it. |
| g5 | Pour the brewed coffee from the French press. | 4:00.1–5:11.8 | sound | audio_helps | X017 X018 X019 | After some background squeaking and footsteps, the wearer pours the brewed liquid from the French press. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:08.0–0:09.0 | speech | Alright. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:34.5–0:35.2 | sound | Gas burner igniting [Gas burner] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:35.2–0:35.5 | sound | Water bubbling in kettle [Kettle] | ours-legacy sound (unverified, model/hand made) |
| X002 | 1:17.3–1:28.9 | sound | hand movement [hands] | ours-legacy sound (unverified, model/hand made) |
| X003 | 1:28.9–1:29.6 | sound | coffee filter falling [coffee filter] | ours-legacy sound (unverified, model/hand made) |
| X004 | 1:30.3–1:31.1 | sound | French press assembly [French press] | ours-legacy sound (unverified, model/hand made) |
| X005 | 1:44.0–1:55.6 | sound | Pouring coffee [coffee] | ours-legacy sound (unverified, model/hand made) |
| X006 | 2:00.4–2:07.2 | sound | opening coffee bag [coffee bag] | ours-legacy sound (unverified, model/hand made) |
| X007 | 2:16.3–2:17.5 | sound | placing kettle on stovetop [kettle] | ours-legacy sound (unverified, model/hand made) |
| X008 | 2:23.1–2:24.2 | sound | taking lid off [kettle] | ours-legacy sound (unverified, model/hand made) |
| X009 | 2:26.4–2:29.9 | sound | putting lid on [kettle] | ours-legacy sound (unverified, model/hand made) |
| X010 | 2:30.4–2:31.0 | sound | Placing [kettle] | ours-legacy sound (unverified, model/hand made) |
| X011 | 2:38.0–2:38.5 | sound | Placing [kettle] | ours-legacy sound (unverified, model/hand made) |
| X012 | 3:28.1–3:28.4 | sound | Gas stove igniting [Gas stove] | ours-legacy sound (unverified, model/hand made) |
| X013 | 3:36.9–3:50.4 | sound | pouring [kettle] | ours-legacy sound (unverified, model/hand made) |
| X014 | 3:37.6–3:49.5 | sound | filling a container | ours-legacy sound (unverified, model/hand made) |
| X015 | 3:52.5–3:52.7 | sound | placing kettle down [kettle] | ours-legacy sound (unverified, model/hand made) |
| X016 | 3:57.2–3:59.7 | sound | placing lid | ours-legacy sound (unverified, model/hand made) |
| X017 | 4:00.1–4:03.2 | sound | Squeaking [Juicer] | ours-legacy sound (unverified, model/hand made) |
| X018 | 4:05.5–5:11.8 | sound | footsteps [person] | ours-legacy sound (unverified, model/hand made) |
| X019 | 4:32.0–4:48.4 | sound | Pouring liquid [french press] | ours-legacy sound (unverified, model/hand made) |
| X020 | 5:06.3–5:09.4 | sound | Typing [phone] | ours-legacy sound (unverified, model/hand made) |
loc3_script4_seq7_rec1
Script 4 · relaxed Sunday morning · 169 s · ASR utterances 11 · legacy sound tags 43
video: SAVVY Video/loc3_script4_seq7_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:reading, games or exercise
overall_goal:Play a game on a mobile phone.
others_involved(模型判断有无他人参与):False
unrelated events:X000, S000, S001, S002, S003, S004, S005, S006, S007, S008, S009, S010 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Navigate the phone interface. | 0:23.6–1:51.5 | sound | sound_only | X001 X002 X003 X004 X005 X006 X007 X008 X009 X010 X011 X012 X013 X014 X015 X016 X017 X018 | Continuous clicking sounds indicate the wearer is interacting with the phone screen to navigate apps or menus. |
| g2 | Play a mobile game. | 1:52.3–2:29.1 | sound | sound_only | X019 X020 X021 X022 X023 X024 X025 X026 X027 X028 X029 X030 X031 X032 X033 X034 X035 X036 X037 X038 X039 X040 | A series of screen clicks followed by distinct game sound effects shows the wearer is actively playing a game. |
| g3 | Lock the phone and put it away. | 2:31.3–2:35.4 | sound | sound_only | X041 X042 | A final click followed by the sound of the phone dropping on a table indicates the wearer has finished playing and set the device down. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| X000 | 0:04.4–0:10.9 | sound | Dog walking [dog] | ours-legacy sound (unverified, model/hand made) |
| S000 | 0:23.0–0:24.0 | speech | ah | official speech.csv (ASR, speaker unknown) |
| X001 | 0:23.6–0:24.0 | sound | Click | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:28.0–0:28.3 | sound | Click | ours-legacy sound (unverified, model/hand made) |
| S001 | 0:44.2–0:46.2 | speech | ah ah ah | official speech.csv (ASR, speaker unknown) |
| S002 | 1:03.9–1:10.8 | speech | ah ah ah ah ah ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S003 | 1:13.5–1:18.2 | speech | ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S004 | 1:30.2–1:31.2 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| S005 | 1:34.1–1:42.9 | speech | ah ah ah ah ah ah ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| X003 | 1:44.9–1:45.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X004 | 1:45.1–1:45.2 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X005 | 1:45.2–1:45.3 | sound | click | ours-legacy sound (unverified, model/hand made) |
| X006 | 1:45.3–1:45.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:46.1–1:46.2 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:46.3–1:46.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X009 | 1:46.6–1:46.7 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X010 | 1:47.7–1:47.9 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X011 | 1:48.0–1:48.1 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X012 | 1:48.2–1:48.3 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X013 | 1:49.3–1:49.4 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X014 | 1:49.5–1:49.7 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X015 | 1:49.8–1:49.9 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X016 | 1:50.8–1:51.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X017 | 1:51.1–1:51.2 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X018 | 1:51.3–1:51.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| S006 | 1:52.3–1:54.6 | speech | ah ah ah | official speech.csv (ASR, speaker unknown) |
| X019 | 1:52.3–1:52.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X020 | 1:52.6–1:52.8 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X021 | 1:52.8–1:53.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X022 | 1:53.9–1:54.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X023 | 1:54.1–1:54.2 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X024 | 1:54.3–1:54.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X025 | 1:55.3–1:55.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X026 | 1:55.6–1:55.8 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X027 | 1:55.8–1:56.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X028 | 1:56.9–1:57.1 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X029 | 1:56.9–1:57.1 | sound | click | ours-legacy sound (unverified, model/hand made) |
| X030 | 1:57.2–1:57.3 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X031 | 1:57.4–1:57.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X032 | 1:57.6–1:57.7 | sound | click | ours-legacy sound (unverified, model/hand made) |
| X033 | 1:58.4–1:58.5 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X034 | 1:58.7–1:58.8 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X035 | 1:58.9–1:59.0 | sound | Click [Phone screen] | ours-legacy sound (unverified, model/hand made) |
| X036 | 2:02.3–2:02.6 | sound | Game sound effect [Phone] | ours-legacy sound (unverified, model/hand made) |
| S007 | 2:04.2–2:05.2 | speech | ah | official speech.csv (ASR, speaker unknown) |
| X037 | 2:08.1–2:08.4 | sound | Game sound effect [Phone] | ours-legacy sound (unverified, model/hand made) |
| X038 | 2:15.7–2:16.0 | sound | Game sound effect [Phone] | ours-legacy sound (unverified, model/hand made) |
| S008 | 2:16.1–2:17.2 | speech | ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| X039 | 2:22.6–2:23.0 | sound | Game sound effect [Phone] | ours-legacy sound (unverified, model/hand made) |
| X040 | 2:28.6–2:29.1 | sound | Game sound effect [Phone] | ours-legacy sound (unverified, model/hand made) |
| X041 | 2:31.3–2:31.6 | sound | Click. [Phone] | ours-legacy sound (unverified, model/hand made) |
| S009 | 2:32.6–2:35.4 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| X042 | 2:34.8–2:35.4 | sound | Phone drops on table. [Phone] | ours-legacy sound (unverified, model/hand made) |
| S010 | 2:47.6–2:48.6 | speech | ah | official speech.csv (ASR, speaker unknown) |
loc3_script5_seq3_rec1
Script 5 · getting home and getting going · 113 s · ASR utterances 1 · legacy sound tags 10
video: SAVVY Video/loc3_script5_seq3_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:indoor-outdoor transition
overall_goal:Remove outerwear or change clothes and check phone
others_involved(模型判断有无他人参与):False
unrelated events:none · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Remove outerwear or change clothes | 0:05.2–1:34.1 | mixed | audio_helps | S000 X000 X001 X002 X003 X004 X005 X006 X007 | The wearer says 'Okay' to initiate the action, followed by continuous clothing rustling sounds indicating they are taking off or changing clothes. |
| g2 | Interact with a mobile phone | 1:43.2–1:52.5 | sound | audio_helps | X008 X009 | Clicking sounds associated with a phone indicate the wearer has picked up and is using their device. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:05.2–0:06.2 | speech | Okay. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:31.3–0:31.7 | sound | Rustling or crinkling sound of clothing, likely due to movement or handling. [clothing] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:33.4–0:34.2 | sound | Rustling or crinkling sound of clothing, likely due to movement or handling. [clothing] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:44.8–0:45.5 | sound | Rustling or crinkling sound of clothing, likely due to movement or handling. [clothing] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:56.8–0:57.1 | sound | Rustling or crinkling sound of clothing, likely due to movement or handling. [clothing] | ours-legacy sound (unverified, model/hand made) |
| X004 | 1:00.0–1:02.4 | sound | Rustling clothes [clothing] | ours-legacy sound (unverified, model/hand made) |
| X005 | 1:15.2–1:22.3 | sound | Rustling clothes [clothing] | ours-legacy sound (unverified, model/hand made) |
| X006 | 1:30.0–1:32.0 | sound | Rustling | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:32.5–1:34.1 | sound | Rustling | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:43.2–1:43.2 | sound | Click [Phone] | ours-legacy sound (unverified, model/hand made) |
| X009 | 1:52.4–1:52.5 | sound | Click [Phone] | ours-legacy sound (unverified, model/hand made) |
loc3_script5_seq6_rec1
Script 5 · getting home and getting going · 158 s · ASR utterances 1 · legacy sound tags 13
video: SAVVY Video/loc3_script5_seq6_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:indoor-outdoor transition
overall_goal:Enter the home, organize clothing into a bag, and check a smartphone.
others_involved(模型判断有无他人参与):False
unrelated events:none · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Enter the room and close the door. | 0:07.7–0:30.7 | mixed | audio_helps | S000 X000 X001 X002 X003 X004 | The wearer says 'Okay' and operates a door, which is evidenced by multiple hinge creaks culminating in a door slam. |
| g2 | Fold clothes and pack them into a bag. | 1:01.5–2:15.0 | sound | audio_helps | X005 X006 X007 X008 X009 | Continuous sounds of clothing being handled and folded are immediately followed by the sound of a bag zipper. |
| g3 | Operate a smartphone. | 2:31.2–2:37.6 | sound | audio_helps | X010 X011 X012 | A series of tapping sounds indicates the wearer is interacting with a phone screen. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:07.7–0:08.7 | speech | Okay. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:11.4–0:12.0 | sound | hinge creak [hinge] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:24.1–0:24.4 | sound | hinge creak [hinge] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:26.6–0:27.2 | sound | hinge creak [hinge] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:28.2–0:28.9 | sound | hinge creak [hinge] | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:30.0–0:30.7 | sound | Door slam [Door] | ours-legacy sound (unverified, model/hand made) |
| X005 | 1:01.5–1:02.9 | sound | Clothing handling [Clothing] | ours-legacy sound (unverified, model/hand made) |
| X006 | 1:07.6–1:24.3 | sound | Clothing handling [Clothing] | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:28.7–1:30.1 | sound | Clothing handling [Clothing] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:30.0–1:56.1 | sound | Clothing folding [clothing] | ours-legacy sound (unverified, model/hand made) |
| X009 | 2:02.0–2:15.0 | sound | zipper [bag] | ours-legacy sound (unverified, model/hand made) |
| X010 | 2:31.2–2:31.3 | sound | Tap [iPhone] | ours-legacy sound (unverified, model/hand made) |
| X011 | 2:35.4–2:35.6 | sound | Tap [iPhone] | ours-legacy sound (unverified, model/hand made) |
| X012 | 2:37.4–2:37.6 | sound | Tap [iPhone] | ours-legacy sound (unverified, model/hand made) |
loc4_script1_seq1_rec1
Script 1 · lazy morning before a party · 165 s · ASR utterances 19 · legacy sound tags 21
video: SAVVY Video/loc4_script1_seq1_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
g4
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:reading, games or exercise
overall_goal:Perform an exercise routine
others_involved(模型判断有无他人参与):False
unrelated events:none · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Prepare the space and equipment for exercise | 0:01.4–0:35.4 | mixed | audio_helps | X000 X001 X002 X003 X004 X005 X006 X007 X008 X009 X010 X011 S000 X012 X013 | Footsteps, object movements, and putting items away (X000-X013) along with an initial breath or grunt (S000) indicate the wearer is moving around and setting up the area before starting the workout. |
| g2 | Perform a set of strenuous exercises | 0:36.6–1:58.5 | speech | audio_helps | S001 S002 S003 S004 S005 S006 S007 S008 S009 S010 | A continuous series of vocalized grunts and heavy breathing captured by ASR (S001-S010) suggests the wearer is actively performing a physically demanding exercise set. |
| g3 | Perform a set of pull-ups | 2:00.7–2:26.7 | mixed | audio_helps | X014 S011 S012 X015 S013 S014 X016 S015 X017 X018 S016 | Machine tags explicitly identifying pull-ups (X014-X018) interspersed with exertion grunts (S011-S016) indicate the wearer is doing pull-up exercises. |
| g4 | Conclude the workout and check the phone | 2:30.3–2:43.8 | mixed | audio_helps | X019 S017 X020 S018 | A clang sound (X019) followed by phone typing (X020) and final resting breaths (S017, S018) suggests the wearer has finished the exercise and is resting while using their phone. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| X000 | 0:01.4–0:02.3 | sound | Footstep [Human] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:03.6–0:05.1 | sound | Hand movement [Human] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:04.3–0:04.8 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:06.8–0:07.0 | sound | Footstep [Human] | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:10.4–0:12.0 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X005 | 0:13.3–0:15.1 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X006 | 0:15.8–0:17.1 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X007 | 0:18.6–0:20.0 | sound | Human respiration [Human] | ours-legacy sound (unverified, model/hand made) |
| X008 | 0:20.9–0:21.2 | sound | Footstep [Human] | ours-legacy sound (unverified, model/hand made) |
| X009 | 0:22.1–0:23.4 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X010 | 0:24.1–0:25.5 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| X011 | 0:26.1–0:27.2 | sound | Object movement | ours-legacy sound (unverified, model/hand made) |
| S000 | 0:26.3–0:27.3 | speech | ah | official speech.csv (ASR, speaker unknown) |
| X012 | 0:29.1–0:30.0 | sound | Hand movement [Human] | ours-legacy sound (unverified, model/hand made) |
| X013 | 0:33.8–0:35.4 | sound | Putting items away | ours-legacy sound (unverified, model/hand made) |
| S001 | 0:36.6–0:37.6 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| S002 | 0:44.1–0:45.1 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| S003 | 0:47.3–0:49.4 | speech | ah ah ah | official speech.csv (ASR, speaker unknown) |
| S004 | 1:05.0–1:07.3 | speech | ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S005 | 1:12.6–1:13.6 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| S006 | 1:18.0–1:22.7 | speech | ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S007 | 1:24.5–1:31.1 | speech | ah ah ah ah ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S008 | 1:39.3–1:43.1 | speech | ah ah ah | official speech.csv (ASR, speaker unknown) |
| S009 | 1:44.4–1:50.7 | speech | ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| S010 | 1:54.1–1:58.5 | speech | ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| X014 | 2:00.7–2:02.3 | sound | pull up [human] | ours-legacy sound (unverified, model/hand made) |
| S011 | 2:02.3–2:03.3 | speech | ah | official speech.csv (ASR, speaker unknown) |
| S012 | 2:09.3–2:11.4 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| X015 | 2:10.6–2:11.0 | sound | pull up [human] | ours-legacy sound (unverified, model/hand made) |
| S013 | 2:13.0–2:14.0 | speech | ah | official speech.csv (ASR, speaker unknown) |
| S014 | 2:16.1–2:17.2 | speech | ah ah | official speech.csv (ASR, speaker unknown) |
| X016 | 2:18.0–2:18.6 | sound | pull up [human] | ours-legacy sound (unverified, model/hand made) |
| S015 | 2:19.4–2:23.0 | speech | ah ah ah ah ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| X017 | 2:20.4–2:21.1 | sound | pull up [human] | ours-legacy sound (unverified, model/hand made) |
| X018 | 2:22.9–2:23.5 | sound | pull up [human] | ours-legacy sound (unverified, model/hand made) |
| S016 | 2:24.2–2:26.7 | speech | ah ah ah ah | official speech.csv (ASR, speaker unknown) |
| X019 | 2:30.3–2:33.7 | sound | Clang | ours-legacy sound (unverified, model/hand made) |
| S017 | 2:31.9–2:32.9 | speech | ah | official speech.csv (ASR, speaker unknown) |
| X020 | 2:39.0–2:41.0 | sound | typing [phone] | ours-legacy sound (unverified, model/hand made) |
| S018 | 2:41.5–2:43.8 | speech | ah ah ah | official speech.csv (ASR, speaker unknown) |
loc4_script1_seq3_rec1
Script 1 · lazy morning before a party · 109 s · ASR utterances 1 · legacy sound tags 9
video: SAVVY Video/loc4_script1_seq3_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:cleaning
overall_goal:Vacuum the room
others_involved(模型判断有无他人参与):False
unrelated events:S000 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Turn on and operate the vacuum cleaner | 0:04.5–1:21.4 | sound | sound_only | X000 X001 X002 X003 X004 X005 X006 | Events X000 through X006 indicate the vacuum cleaner being switched on and continuously operated. |
| g2 | Stop vacuuming and tidy up | 1:36.4–1:47.7 | sound | audio_helps | X007 X008 | A thump from the vacuum in X007 suggests it is turned off or put away, followed by the sound of placing a book in X008. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| S000 | 0:00.5–0:02.7 | speech | Alright, I'm recording again. | official speech.csv (ASR, speaker unknown) |
| X000 | 0:04.5–0:05.4 | sound | vacuum cleaner switch on [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:07.7–0:30.1 | sound | vacuum cleaning [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:31.8–0:32.5 | sound | Vacuum cleaner operating [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:49.7–0:59.2 | sound | Vacuum cleaner operating [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X004 | 1:00.7–1:01.7 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X005 | 1:13.7–1:16.8 | sound | vacuum cleaner [vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X006 | 1:20.1–1:21.4 | sound | electrical sounds | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:36.4–1:36.5 | sound | Thump [Vacuum cleaner] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:46.7–1:47.7 | sound | Book placing | ours-legacy sound (unverified, model/hand made) |
loc4_script5_seq3_rec1
Script 5 · getting home and getting going · 159 s · ASR utterances 1 · legacy sound tags 14
video: SAVVY Video/loc4_script5_seq3_rec1.mp4(20 fps 去畸变,mic 5+6);时间 0 = 第一帧 RGB
g1
g2
g3
g4
上条 sub-goal 区段;下条事件刻度(绿 = ASR 语音,灰 = legacy sound)
Goal / sub-goal 草稿(Gemini 3.1 Pro,仅 ASR + sound 事件,未核验)
activity_type:organization and laundry
overall_goal:Organize clothes, handle a suitcase, and read a book
others_involved(模型判断有无他人参与):False
unrelated events:S000, X013 · validation errors 0
| # | sub-goal | clip time | completion | audio_necessity | events | rationale |
|---|---|---|---|---|---|---|
| g1 | Make the bed and fold clothes | 0:03.8–0:40.2 | sound | audio_helps | X000 X001 X002 X003 X004 | Sounds of bed making, folding clothes, and rustling indicate the wearer is organizing items in the bedroom. |
| g2 | Open a door or closet to handle clothes | 0:40.7–1:01.6 | sound | audio_helps | X005 X006 X007 | A hinge squeak followed by rustling clothes and footsteps suggests opening a closet or room door to put away or retrieve garments. |
| g3 | Pack or unpack a suitcase | 1:31.8–1:53.5 | sound | audio_helps | X008 X009 | Rustling clothing sounds combined with suitcase sounds indicate the wearer is interacting with luggage. |
| g4 | Read or handle a book | 1:55.2–2:17.6 | sound | audio_helps | X010 X011 X012 | Page turning and general book sounds show the wearer has transitioned to reading or looking through a book. |
事件时间线
| id | clip time | kind | text | source |
|---|---|---|---|---|
| X000 | 0:03.8–0:11.8 | sound | Bed making. [Bed] | ours-legacy sound (unverified, model/hand made) |
| X001 | 0:14.7–0:31.5 | sound | Folding clothes [Clothing] | ours-legacy sound (unverified, model/hand made) |
| X002 | 0:30.8–0:32.1 | sound | Rustling [Clothes] | ours-legacy sound (unverified, model/hand made) |
| X003 | 0:33.4–0:36.2 | sound | Rustling [Clothes] | ours-legacy sound (unverified, model/hand made) |
| X004 | 0:37.9–0:40.2 | sound | Rustling [Clothes] | ours-legacy sound (unverified, model/hand made) |
| X005 | 0:40.7–0:42.1 | sound | Hinge Squeak [Door] | ours-legacy sound (unverified, model/hand made) |
| S000 | 0:44.3–0:45.3 | speech | you. | official speech.csv (ASR, speaker unknown) |
| X006 | 0:44.9–0:58.5 | sound | Rustling [Clothes] | ours-legacy sound (unverified, model/hand made) |
| X007 | 1:00.5–1:01.6 | sound | Footstep [feet] | ours-legacy sound (unverified, model/hand made) |
| X008 | 1:31.8–1:42.4 | sound | Rustling sounds [Clothing] | ours-legacy sound (unverified, model/hand made) |
| X009 | 1:43.3–1:53.5 | sound | Suitcase [Suitcase] | ours-legacy sound (unverified, model/hand made) |
| X010 | 1:55.2–1:59.5 | sound | Page Turn [Book] | ours-legacy sound (unverified, model/hand made) |
| X011 | 2:00.0–2:07.8 | sound | Book sounds [Book] | ours-legacy sound (unverified, model/hand made) |
| X012 | 2:16.3–2:17.6 | sound | Book sounds [Book] | ours-legacy sound (unverified, model/hand made) |
| X013 | 2:38.0–2:38.4 | sound | finger snap | ours-legacy sound (unverified, model/hand made) |