GOP and LPR score

Recently, I find 2 parts are different from the original Hu's GOP paper.

1. [LPR computation](https://github.com/kaldi-asr/kaldi/blob/71f38e62cad01c3078555bfe78d0f3a527422d75/src/bin/compute-gop.cc#LL230C1-L230C1)

```C++
// LPR(p_j|p_i)=\log p(p_j|\mathbf o; t_s, t_e)-\log p(p_i|\mathbf o; t_s, t_e)
for (int k = 0; k < phone_num; k++)
phone_level_feat(1 + phone_num + k) = lpp_part(phone_id) - lpp_part(k);
```

Per my understanding, $p_i$ should be the canonical phoneme, $LPR(p_j|p_i) = \log p(p_j|\mathbf o; t_s, t_e) - \log p(p_i|\mathbf o; t_s, t_e)$, and phoneme level feature is defined as ${[LPP(p_1),\cdots,LPP(p_M), LPR(p_1|p_i), \cdots, LPR(p_j|p_i),\cdots]}^T$. So I think the above code should be changed as:

```C++
// LPR(p_j|p_i)=\log p(p_j|\mathbf o; t_s, t_e)-\log p(p_i|\mathbf o; t_s, t_e)
for (int k = 0; k < phone_num; k++)
phone_level_feat(1 + phone_num + k) = pp_part(k) - lpp_part(phone_id);
```

2. Formulation to compute the GOP score in the document

In the document,

$$GOP(p)=\log \frac{LPP(p)}{\max_{q\in Q} LPP(q)}$$

$$LPP(p)=\log p(p|\mathbf o; t_s,t_e)$$

In Hu's paper

$$GOP(p)=\log \frac{p(p|\mathbf o; t_s,t_e)}{\max_{q\in Q} p(q|\mathbf o; t_s,t_e)}$$

Thus, I think the GOP formulation in the document should be changed to

$$GOP(p)=\frac{LPP(p)}{\max_{q\in Q} LPP(q)}$$

Start with src/bin/compute-gop.cc at the linked LPR computation, then locate the GOP formulation in the project documentation. Compare both implementations and formulas with Hu's paper, checking the canonical-phoneme reference and logarithm relationships. Done means the documented equations and implementation consistently match the intended GOP and LPR definitions.

Written by the indexing model from the issue text.