Mod: Slight code cleaning of CIDEr-D computation.

src/aac_metrics/functional/cider_d.py

@@ -262,25 +262,21 @@ def __compute_cider(
         # compute vector for test captions
         vec, norm, length = __counter_to_vec(cand, log_n_refs, n, doc_frequencies)
         # compute vector for ref captions
-        ngrams_scores = np.zeros((n,))
+        ngrams_scores = np.zeros((len(refs), n))
         vec_refs = []
-        for ref in refs:
+        for j, ref in enumerate(refs):
             vec_ref, norm_ref, length_ref = __counter_to_vec(
                 ref, log_n_refs, n, doc_frequencies
             )
             vec_refs.append(vec_ref)
-            ngrams_scores += __similarity(
+            ngrams_scores[j] = __similarity(
                 vec, vec_ref, norm, norm_ref, length, length_ref, n, sigma
             )
-        # change by vrama91 - mean of ngram scores, instead of sum
-        score_avg = np.mean(ngrams_scores)
-        # divide by number of references
-        score_avg /= len(refs)
-        # multiply score by 10
-        score_avg *= scale
-        # append score of an image to the score list
-        scores[i] = score_avg
 
+        # Use this weird mean calculation instead of ".mean()" because it can give slight differences compared to the original implementation
+        score_avg = ngrams_scores.sum(axis=0).mean() / len(refs)
+        scores[i] = score_avg
         tfidf_lst.append((vec, vec_refs))
 
+    scores = scores * scale
     return scores, tfidf_lst