Processing Blocks

1. argmax

This proc-block returns indices of the max element of the array.

• Types
  • Supported input types - f32
  • Supported output types - u32

2. audio_float_conversion

Many audio models take in an input of f32. This proc-block convert our input from an i16 data type to a floating-point value.

• Types
  • Supported input types - i16
  • Supported output types - f32

3. binary_classification

A proc-block takes a probability (0.0 to 1.0) score as input and divides the output into two classes ( 0 or 1) based on a threshold. It returns 1 if the probability value is greater than the threshold otherwise, 0. This is useful in binary classification problems.

• Types
  • Supported input types - f32
  • Supported output types - u32
• args
  • threshold - criteria to divide output into classes

4. fft

This proc-block converts a signal from its original domain (often time or space) to a representation in the frequency domain.

• Types
  • Supported input types - i16
  • Supported output types - u32
• args
  • Sample Rate - Sampling rate
  • Bins - intervals between samples in frequency domain
  • Window Overlap - Ratio of overlapped intervals.

5. image-normalization

A normalization routine takes the image matrix as input and fits their values to the range [0, 1] as f32's.

• Types
  • Supported input types - u8, u16, u32, i8, i16, i32
  • Supported output types - f32

6. label

A proc block, when given a set of indices, will return their associated labels.

• Types
  • Supported input types - u8, u16, u32, u64, i8, i16, i32, i64, f32, f64
  • Supported output types - utf8
• args
  • wordlist - Upload the wordlist in the .txt format with every label in different line.

7. modulo

As the same suggests, it returns the remainder of a division after one number is divided by another.

• Types
  • Supported input types - u8, u16, u32, u64, i8, i16, i32, i64, f32, f64
  • Supported output types - utf8

8. most_confident_indices.

A proc block which, when given a list of confidences, will return the indices of the top N most confident values.

• Types
  • Supported input types - u8, u16, u32, u64, i8, i16, i32, i64, f32, f64
  • Supported output types - u32
• args
  • Count - number of classes with highest confidence you want as output

9. noise-filtering

This proc-block perform a couple of functions:

• Reduces noise within each frequency bin (channel)
• Applies a gain control algorithm to each frequency bin (channel)
• Applies log2 function and scales the output.

It reduces noise and applies a gain control algorithm within each frequency bin.

• Types
  • Supported input types - u32
  • Supported output types - i16

10. normalize

This proc-block normalizes the input to the range [0, 1].

• Types
  • Supported input types - u8, u16, u32, u64, i8, i16, i32, i64, f32, f64
  • Supported output types - f32
• args
  • Count - number of classes with highest confidence you want as output

11. object_filter

A proc-block which takes 3-d tensor [1, num_detection, detection_box(x, y, w, h) + confidence_scores + total_detection_classes] and filter the detected objects to:

• remove duplicate detection for a single object
• remove the objects with low confidence based on a threshold

giving a 2-d tensor with dimension [*, 6] (where * is the total number of detected objects and 6 -> [ x-coordinate, y-coordinate, h, w, confidence_value, label_index]) as output. It is used in object detection models.

• Types
  • Supported input types - f32
  • Supported output types - f32
• args
  • Threshold - remove the objects with confidence value below this threshold

12. parse

A proc block that can parse a string to numbers. This proc-block could be helpful in doing non-ML tasks.

• Types
  • Supported input types - utf8
  • Supported output types - u8, u16, u32, u64, i8, i16, i32, i64, f32, f64

13. segment_output

A proc-block which takes a rank 4 tensor as input, whose dimension is of this form [1, x, y, z]. It will return:

• a 2-d tensor after performing argmax along the axis-3 of the tensor

• a 1-d tensor which a set of all the numbers present in the above 2-d tensor

  • Types
    • Supported input types - f32
    • Supported output types - u32

14. softmax

This proc-block returns tensor after applying the softmax function over the array.

• Types
  • Supported input types - f32
  • Supported output types - f32

15. tokenizer

A proc-block takes a passage and a question as input and gives us input_ids, input_masks, segment_ids as output which are then fed to the model. This is helpful in NLP models.

• Types
  • Supported input types
    • input-1 types - u8
    • input-2 types - u8
  • Supported output types
    • output-1 types - i32
    • output-2 types - i32
    • output-3 types - i32
    • output-3 types - u8

16. text_extractor

A proc-block takes start logits, end logits, and ASCII coded text and returns an output sentence (utf8 encoded) from start logit to ends logit.

• Types
  • Supported input types
    • input-1 types - u8
    • input-2 types - u32
    • input-3 types - u32
  • Supported output types
    • output-1 types - utf8