RESEARCH

Acoustic & Speech

Investigates various Speech signal processing schemes for acoustic modeling so that more robust speech recognition can be achieved. Our aim is to perform the state-of-art research providing effective means for achieving:

Speech Recognizer on Fixed-point Processor

Admin    2021-04-08    👁️ 498

Speech Recognizer on Fixed-point Processor
　

     Contents

1. Fixed-Point Issues

2. Design principles for fixed Q-point variables

3. Software Design Issues

4. Application : Speech Recognizer on PDA

1. Fixed-Point Issues         

• The algorithm has to carefully normalize its numerical behavior to within the dynamic range (data format limitation) of a 16/32-bit integer at every stage of the computation.

• Fixed-Point DSP processor usually incorporates a hardware multiplier so that addition and multiplication can be completed in one CPU cycle. However, there is no hardware for division. So for doing this, it takes more than 20 cycles by a routine.
  ( Explicit division : pre-compute the inverted data. Division can be approximated by multiplication and shift. )

• Implicit division : pointer arithmetic is used heavily in the memory management in the search algorithm. Pointer subtraction actually incurs a division.

2. Design principles for fixed Q-point variables         

• Compute dynamic ranges for all variables.

• Inspect the dynamic range and choose a fixed Q-point for each variable.

• For array variables ( ex. MFCC ), each dimension may have different dynamic range. Choose a normalization factor for each of the dimension to normalize the dynamic range so they can all use the same Q-point.

• Choose an appropriate place in the algorithm to incorporate the normalization factors.

3. Software Design Issues        

• Streamline data structures so that all model data are efficient to store and also computable, as opposed to the typical two format scheme: one for disk storage and the other for computation.

• Examine the allocation of local and global variables to balance memory efficiency and program modularity. Local variables consume stack size. This is especially important for recursive routines.

• Replace all string operations with efficient integer operations.

• Remove all malloc() and free(). Design algorithm to do memory management and garbage collection. The algorithm is tailored for exact data structure to be efficient.

4. Application : Speech Recognizer on PDA