Subject: Re: [AUDITORY] Visualizing Auditory Network Embedding of English Words From: Roger K Moore <0000011559506d60-dmarc-request@xxxxxxxx> Date: Mon, 5 Oct 2026 07:44:43 +0100--Apple-Mail-7BE73771-1441-4825-8FB5-0066408994A5 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html class=3D"apple-mail-supports-explicit-dark-mode"><head><meta http-equi= v=3D"content-type" content=3D"text/html; charset=3Dutf-8"></head><body dir=3D= "auto">Interesting =E2=80=A6 I was struck how non-intuitive the phonetic sim= ilarities were. It might be fun to compare with the word confusion mod= el I developed in the 1970s! <a href=3D"https://ieeexplore.ieee.org/do= cument/1162916">https://ieeexplore.ieee.org/document/1162916</a><br id=3D"li= neBreakAtBeginningOfSignature"><div dir=3D"ltr"><br></div><div dir=3D"ltr">R= oger</div><div dir=3D"ltr"><br><blockquote type=3D"cite">On 5 Oct 2026, at 0= 5:55, Malcolm Slaney <000001757ffb5fe1-dmarc-request@xxxxxxxx> w= rote:<br><br></blockquote></div><blockquote type=3D"cite"><div dir=3D"ltr">=EF= =BB=BF<meta http-equiv=3D"content-type" content=3D"text/html; charset=3Dutf-= 8"><meta http-equiv=3D"content-type" content=3D"text/html; charset=3Dus-asci= i"><div style=3D"overflow-wrap: break-word; -webkit-nbsp-mode: space; line-b= reak: after-white-space;">Why are some words easier to recognize than others= ? Some talk about perceptual neighborhood density, a measure of how many wor= ds sound similar to a query word. Visualizing these distances is the first s= tep. Thus...<div><br></div><div>Check out the acoustic network embedding (AN= E) visualization at</div><div><span class=3D"Apple-tab-span" style=3D"white-= space:pre"> </span>https://quicksin.stanford.edu/ane</div><div><br></di= v><div><span style=3D"font-family: Arial, Verdana, sans-serif; font-variant-= ligatures: normal; background-color: rgb(255, 255, 255);">Enter an English w= ord in the search bar and see the words that are closest to your chosen word= . You can change the number of neighbors displayed and the underling e= mbedding dimension. Move your mouse over a word to see it=E2=80=99s distance= from the search word.</span></div><div><span style=3D"font-family: Arial, V= erdana, sans-serif; font-variant-ligatures: normal; background-color: rgb(25= 5, 255, 255);"><br></span></div><div><span style=3D"font-family: Arial, Verd= ana, sans-serif; font-variant-ligatures: normal; background-color: rgb(255, 2= 55, 255);">This visualization is based on Woojay Jeon's (@xxxxxxxx)</span><span= style=3D"font-family: Arial, Verdana, sans-serif; font-variant-ligatures: n= ormal; background-color: rgb(255, 255, 255);"> </span><a href=3D"https:= //github.com/apple-aiml-research/ml-acn-embed" _fcksavedurl=3D"https://githu= b.com/apple-aiml-research/ml-acn-embed" style=3D"font-family: Arial, Verdana= , sans-serif; font-variant-ligatures: normal;">Acoustic Neighbor Embeddings (= ACN)</a>. His network learns a high-dimensional embedding that puts each Eng= lish word into a point in a multi-dimensional space so that acoustic distanc= es between words are preserved.</div><div><br></div><div>For more details se= e this blog posting: https://ccrma.stanford.edu/blogs/malcolm/visualizi= ng-word-acoustic-neighborhood-embeddings-ane</div><div><br></div><div>Have f= un! Go to https://quicksin.stanford.edu/ane</div><div><br></div><div>= - Malcolm</div><div><br></div></div></div></blockquote></body></html>= --Apple-Mail-7BE73771-1441-4825-8FB5-0066408994A5--