Loading of new TIMIT data added : formant values and bandwidth annotations

src/corpora/timit.jl

@@ -224,9 +224,12 @@ TIMIT_PHONE_MAP39 = Dict(
 #######################################################################
 
 
-function timit_prepare(timitdir, dir; audio_fmt="SPHERE")
+function timit_prepare(timitdir, dir, formantsdir=nothing; audio_fmt="SPHERE")
     # Validate the data directory
     ! isdir(timitdir) && throw(ArgumentError("invalid path $(timitdir)"))
+    if formantsdir !== nothing
+        ! isdir(timitdir) && throw(ArgumentError("invalid path $(timitdir)"))
+    end
 
     # Create the output directory.
     dir = mkpath(dir)
@@ -256,9 +259,9 @@ function timit_prepare(timitdir, dir; audio_fmt="SPHERE")
 
     # Annotations
     @info "Extracting annotations from $timitdir/train"
-    train_annotations = timit_annotations(joinpath(timitdir, "train"))
+    train_annotations = timit_annotations(joinpath(timitdir, "train"), formantsdir)
     @info "Extracting annotations from $timitdir/test"
-    test_annotations = timit_annotations(joinpath(timitdir, "test"))
+    test_annotations = timit_annotations(joinpath(timitdir, "test"), formantsdir)
     annotations = merge(train_annotations, test_annotations)
 
 
@@ -332,8 +335,15 @@ function timit_recordings(dir::AbstractString; fmt="SPHERE")
 end
 
 
-function timit_annotations(dir)
+function timit_annotations(dir, formantsdir=nothing)
     ! isdir(dir) && throw(ArgumentError("expected directory $dir"))
+    
+    if formantsdir !== nothing
+        ddir = last(splitdir(dir))
+        formantsdir = joinpath(formantsdir, ddir)
+        ! isdir(formantsdir) && throw(ArgumentError("expected directory $formantsdir"))
+    end
+
     splitline(line) = rsplit(line, limit=3)
 
     annotations = Dict()
@@ -349,17 +359,38 @@ function timit_annotations(dir)
             (idtuple in processed) && continue
             push!(processed, (dialect, spk, name))
 
-            # Words
+            # Word
             wpath = joinpath(root, name * ".wrd")
             words = [last(split(line)) for line in eachline(wpath)]
 
             # Phones
-            ppath = joinpath(root, name * ".phn")
-            palign = Tuple{Int,Int,String}[]
-            for line in eachline(ppath)
-                t0, t1, p = split(line)
-                push!(palign, (parse(Int, t0), parse(Int, t1), String(p)))
+            ppath = joinpath(root, name * ".phn")            
+            if formantsdir !== nothing
+                forpath = joinpath(formantsdir, dialect, spk, name * ".ft")
+            else 
+                forpath = ""
+            end
+            
+            if isfile(forpath)
+                # Read availabled formants values  
+                palign = Tuple{Int,Int,String,NTuple{8, Float32}}[]
+                for line in eachline(forpath)
+                    t0, t1, p, f1, f2, f3, f4, b1, b2, b3, b4 = split(line)
+                    push!(palign, 
+                    (
+                        parse(Int, t0), parse(Int, t1), String(p),
+                        (parse(Float32, f1), parse(Float32, f2), parse(Float32, f3), parse(Float32, f4),
+                        parse(Float32, b1), parse(Float32, b2), parse(Float32, b3), parse(Float32, b4)))
+                    )
+                end
+            else 
+                palign = Tuple{Int,Int,String}[]
+                for line in eachline(ppath)
+                    t0, t1, p = split(line)
+                    push!(palign, (parse(Int, t0), parse(Int, t1), String(p)))
+                end
             end
+            
 
             sentence_type = if startswith(name, "sa")
                 "dialect"
@@ -391,12 +422,12 @@ function timit_annotations(dir)
 end
 
 
-function TIMIT(timitdir, dir, subset)
+function TIMIT(timitdir, dir, subset, formantsdir=nothing)
     if ! (isfile(joinpath(dir, "recordings.jsonl")) &&
           isfile(joinpath(dir, "annotations-train.jsonl")) &&
           isfile(joinpath(dir, "annotations-dev.jsonl")) &&
           isfile(joinpath(dir, "annotations-test.jsonl")))
-        timit_prepare(timitdir, dir)
+        timit_prepare(timitdir, dir, formantsdir)
     end
     dataset(dir, subset)
 end