Pablo Alonso

Results 10 issues of Pablo Alonso

Some TensorflowPredict derivate algorithms (e.g., TensorflowPredictMusiCNN) have redundant unit test that are not required. In such cases, we could use dummy tests (e.g., `testDummyInvalidParam`) covering redundant tests, with a comment...

enhancement
TensorFlow

One of the goals of Essentia is to maximize parametrizability to allow compatibility with relevant implementations from other frameworks. The following feature requests for [Windowing](https://essentia.upf.edu/reference/std_Windowing.html) allow replicating some framing/windowing options...

enhancement

As suggested in #1174, this feature should be tested. Consider the following cases: - Essentia arrays containing only NaNs - Essentia arrays with mixed NaNs as float23 values - `Yaml`,...

- [ ] TensorFlowPredictOpenL3 - [ ] TensorFlowPredictSpleeter

algorithms wishlist
TensorFlow

**Describe the bug** Our objective is to port an EfficientNet trained in PyTorch and exported in ONNX to TensorFlow. However, we are facing the problem of slower inference times in...

**Describe the bug** I have an ONNX model (opset 11) exported from PyTorch's [EffficientNet](https://github.com/lukemelas/EfficientNet-PyTorch) implementation. My goal is to convert the ONNX model to TensorFLow (SavelModel) to perform inference from...

[TensorToVectorReal](https://essentia.upf.edu/reference/streaming_TensorToVectorReal.html) converts tensors to 2D arrays by flattening all axis but the last one into the first dimension. model-specific prediction algorithms (e.g., [TensorflowPredictVGGish](https://essentia.upf.edu/reference/std_TensorflowPredictVGGish.html)) use this algorithm to return 2D arrays...

Gaia generates empty points when parsing JSON files considered valid YAMLs with the [C++ YAML loader](https://github.com/MTG/gaia/blob/master/src/yamlcpp.cpp). These JSONs are considered valid YAMLs by several online parsers such as [yamllint](http://www.yamllint.com/) or...

Some unit tests have errors or are failing and problems from #3 are still present. The following log was obtaining from Gaia compiled on Ubuntu 18.04 LTS from commit e595e1df9ac6a4444297ef0d87cd34e95e1c4f73...

enhancement

TODO: - [ ] Read `n_classes` from the model parameters.