Skip to content

[speechx]add linear spectrogram feature extractor #1400

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 9 commits into from
Feb 25, 2022

Conversation

SmileGoat
Copy link
Contributor

PR types

[New features]

PR changes

Describe

add linear feature extractor

@mergify mergify bot added the Deployment label Jan 27, 2022
@zh794390558 zh794390558 added this to the r0.2.0 milestone Jan 27, 2022
@zh794390558 zh794390558 mentioned this pull request Jan 27, 2022
6 tasks
@SmileGoat SmileGoat changed the title [WIP][speechx]add linear spectrogram feature extractor, test=doc [WIP][speechx]add linear spectrogram feature extractor Jan 27, 2022
// limitations under the License.

// wrap the fbank feat of kaldi, todo (SmileGoat)

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

header guard

public:
explicit FbankExtractor(const FbankOptions& opts,
share_ptr<FeatureExtractorInterface> pre_extractor);
virtual void AcceptWaveform(const kaldi::Vector<kaldi::BaseFloat>& input) = 0;
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

input -> wave


class FeatureExtractorInterface {
public:
virtual void AcceptWaveform(const kaldi::Vector<kaldi::BaseFloat>& input) = 0;
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

input -> wave

// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

header guard

@@ -0,0 +1,39 @@

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

后面看下是否能放到 speehx/test 里,作为单侧


void SplitFeature(kaldi::Matrix<BaseFloat> feature,
int32 chunk_size,
std::vector<kaldi::Matrix<BaseFloat>> feature_chunks) {
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

std::vector<kaldi::Matrix> *

string utt = feature_reader.Key();
const kaldi::Matrix<BaseFloat> feature = feature_reader.Value();
vector<Matrix<BaseFloat>> feature_chunks;
SplitFeature(feature, chunk_size, &feature_chunks);
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

decoder.InitDecoder(); 放这里

#include "base/basic_types.h"

struct DecoderResult {
BaseFloat acoustic_score;
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

解码器的结果是 score=AMScore+LMScore+...
这个应该是解码器的total分数吧?

@zh794390558 zh794390558 changed the base branch from develop to speechx February 25, 2022 10:26
@zh794390558 zh794390558 changed the title [WIP][speechx]add linear spectrogram feature extractor [speechx]add linear spectrogram feature extractor Feb 25, 2022
@zh794390558 zh794390558 merged commit b584b96 into PaddlePaddle:speechx Feb 25, 2022
@SmileGoat SmileGoat deleted the feature_dev branch March 15, 2022 01:33
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants