CarolinePascal
e2f3982e2c
style(config validation): storing microphone config validation in dedicated methods
2026-01-20 12:20:27 +01:00
CarolinePascal
95de732e55
[skip-ci] feat(multiprocess audio stream): running input audio stream in a separate process to avoid extensive CPU usage impacts
2026-01-20 12:20:27 +01:00
CarolinePascal
b2383236ca
[skip-ci] fix(audio buffers): fill audio buffers before entering the control loop to avoid first audio chunk reading delay
2026-01-20 12:20:27 +01:00
CarolinePascal
4b98cc25c8
[skip-ci] fix(async read): remove async read from LeKiwi to avoid delays
2026-01-20 12:20:27 +01:00
CarolinePascal
90780c4de8
[skip-ci] feat(portaudio timeout): decresing portaudio stream timeout for low latency
2026-01-20 12:20:27 +01:00
CarolinePascal
6f6e046c53
[skip-ci] feat(audio rerun): adding support for multiple channels visualization with rerun
2026-01-20 12:20:27 +01:00
CarolinePascal
8cd64eaad1
fix(resnet weights): fix default ResNet weights for audio
2026-01-20 12:20:27 +01:00
CarolinePascal
e620395416
feat(init audio buffer): adding a separate parameter for the initial audio buffer size
2026-01-20 12:20:27 +01:00
CarolinePascal
0fbcbcdb2e
feat(audio ACT): removing normalization and pretrained weights because it does not really make sense
2026-01-20 12:20:26 +01:00
CarolinePascal
7d430c8067
style(return types): adding missing return types
2026-01-20 12:20:26 +01:00
CarolinePascal
5f114c1d74
feat(init audio buffers): adding correct audio buffer initialization with actually recorded background noise instead of pure silence
2026-01-20 12:20:26 +01:00
CarolinePascal
ad01ef19f4
fix(audio buffers): add security crop to avoid audio buffer overfilling
2026-01-20 12:20:26 +01:00
CarolinePascal
97e91698fb
[skip-ci] fix(online audio chunks): adding missing observation copy when adding audio buffer
2026-01-20 12:20:26 +01:00
CarolinePascal
af0294198a
[skip ci] feat(visualize audio): adding audio recordings visualization in rerun
2026-01-20 12:20:22 +01:00
CarolinePascal
421fdcce96
[skip ci] fix(audio latency): setting microphone recording latency to low as default
2026-01-20 12:12:59 +01:00
CarolinePascal
bb63ad9715
fix(online audio chunks): querying audio chunks as a shifting widow over previous audio samples to match the default audio chunk size
2026-01-20 12:12:59 +01:00
CarolinePascal
3c90a79c57
feat(audio in ACT): adding audio features support in ACT using mel-spectrogram representation
2026-01-20 12:12:59 +01:00
CarolinePascal
8e29c530ed
fix(pytorch audio format): switching to pytorch's default channel first format for audio
2026-01-20 12:12:59 +01:00
CarolinePascal
b573b7a052
fix(audio decoding): fixing edge cases where the requested audio chunk starts before the beginning of the recording
2026-01-20 12:12:59 +01:00
CarolinePascal
926184110b
feat(audio in policies): adding audio as a input feature in policies
2026-01-20 12:12:59 +01:00
CarolinePascal
f73db4394b
fix(audio chunks): querying audio chunks in the past rather than in the future
2026-01-20 12:12:59 +01:00
CarolinePascal
bff91f9927
feat(torchcodec): setting torchcodec as default as the new official release supports audio decoding
2026-01-20 12:12:59 +01:00
CarolinePascal
6d726266fd
fix(audio load file): adding missing dimension when loading mono audio data
2026-01-20 12:12:59 +01:00
CarolinePascal
2962330bb1
style(imports): simplify soundfile imports
2026-01-20 12:12:59 +01:00
CarolinePascal
067993bb11
fix(typos): fixing typos
2026-01-20 12:12:58 +01:00
CarolinePascal
e4dd00c8f5
fix(audio feature shape): fixing audio feature shape ordering (frames first, channels second)
2026-01-20 12:12:58 +01:00
CarolinePascal
3bbd161cfd
[skip ci] feat(audio recording): adding new asyn start_recording, stop_recording and read functions to avoid for loop delays
2026-01-20 12:12:58 +01:00
CarolinePascal
6d7be63f59
[skip ci] fix(wav format): fix default WAV format for audio storage
2026-01-20 12:12:58 +01:00
CarolinePascal
b9d0dfb9a2
[skip ci] refactor(Microphone class): properly rename Microphone class to PortAudioMicrophone to avoid conflicts with the eponym base class
2026-01-20 12:12:58 +01:00
CarolinePascal
dce483060f
[skip ci] feat(audio recording): handle folder creation in start_recording directly
2026-01-20 12:12:58 +01:00
CarolinePascal
c32b9182d9
[skip ci] feat(torchcodec): adding support for torchcodec audio decoding
2026-01-20 12:12:58 +01:00
CarolinePascal
9a5c96b2b1
fix: fixing issue with microphone channels numbering and status recovery on stop_recording
2026-01-20 12:12:58 +01:00
CarolinePascal
688195fc46
docs: add methods descriptions and comments on tricky parts
2026-01-20 12:12:58 +01:00
CarolinePascal
99eb0bbafc
Adding last missing audio features in LeRobotDataset
2026-01-20 12:12:58 +01:00
CarolinePascal
16de8b3f19
Adding support for audio data recording and broadcasting for LeKiwi
2026-01-20 12:12:55 +01:00
CarolinePascal
580008663b
Adding flag for file writting recording case
2026-01-20 12:12:07 +01:00
CarolinePascal
52c424c5eb
Adding multiprocessing support for audio recording
2026-01-20 12:12:07 +01:00
CarolinePascal
836195e59c
Renamming sampling rate to sample rate for consistency
2026-01-20 12:12:07 +01:00
CarolinePascal
be09a59e05
Adding audio tests
2026-01-20 12:12:03 +01:00
CarolinePascal
373a169bd2
Fixing sounddevice stream active state recovery and adding corresponding exceptions
2026-01-20 12:10:45 +01:00
CarolinePascal
00536c6c5b
Adding missing features for audio frames verification and stats
2026-01-20 12:10:45 +01:00
CarolinePascal
cdd3a859ef
Adding pytorch compatible conversion for audio
2026-01-20 12:10:45 +01:00
CarolinePascal
5276fc0d6f
Remove variable audio recordings data types (will be converted to float32 anyway)
2026-01-20 12:10:45 +01:00
CarolinePascal
6a2882f978
Adding audio frames reading capability
2026-01-20 12:10:37 +01:00
CarolinePascal
8874547353
Adding microphone recording in control loop
2026-01-20 12:10:12 +01:00
CarolinePascal
2864caad80
Adding audio modality in LeRobotDatasets
2026-01-20 12:10:12 +01:00
CarolinePascal
d998660aa1
Integrate microphones in Robot class
2026-01-20 12:10:09 +01:00
CarolinePascal
7e5f3b35e9
Implementing basic integration of microphones using soundfile and sounddevice
2026-01-20 12:08:27 +01:00
Alexis D
13bfee1aa4
Set 10 direction bit for Current Load attribute ( #1014 )
2026-01-20 11:20:30 +01:00
Jade Choghari
79688a09f2
improve(dataset-tools): image2video editing tools : Multiple episodes per video file ( #2811 )
...
* improve image2video
* add episodes video encoding
* fix mypy failing
* iterate on review
* nit
* remove max, and let it be optional
* iterate more
* update docs
* fix test
---------
Co-authored-by: Michel Aractingi <michel.aractingi@huggingface.co >
2026-01-20 11:04:22 +01:00
